Zhaoxin x86-Prozessoren

[Benutzername]

Ist der vertreib des HP PC international?

The website is IN-EN, so india english. So I guess HP is marketing towards developing world customers. And as an "internet typewriter" it is fast enough. Though price wise if you have the choice an AMD APU or some intel NUC or something like that is still the better buy.


In der zusammenfassung des THG test wird er ja ganz schön auseinandergenommen.

Yeah, well they compare it to contemporary CPUs from intel and AMD which all are much smaller nodes. Of course it is going to lose in this comparison.

A fairer comparison would have been imho to compare it to older CPUs of a similar process node to see where Zhaoxin stands at the moment and how it compares to where AMD and intel have been back then.

[Tralalak]

PConline: HP 268 Pro G1 review
source: https://translate.google.com/translate?hl=sk&sl=zh-CN&tl=en&u=https%3A%2F%2Fdesktops.pconline.com.cn%2F1335%2F13354004.html

[Tralalak]

HP 268 Pro G1 mainboard with Zhaoxin KX-U6780A processor

8GkoocssCKk


v61ixkPTeJQ

[Loeschzwerg]

I like those board layouts OEM manufacturers do. Simple, clean and no "bling-bling" gamer stuff or color schemes.

[davidzo]

color schemes. I like that HP boards color Scheme. But they could have used different colors, hatching or ENIG surface to distinguish the different screw types for the board, heatsink, etc.

[Benutzername]

Why is the video in english? For India, Africa, Philippines and such places? I doubt this computer is being sold much in the USA.


----------

speaking of USA, Gamers Nexus has posted further review of their THTF TZ561-V3 prebuilt:

RIgBsz1MduI

Also explaining the various operating systems, that are marketed under the KylinOS name.

[Tralalak]

Zhaoxin roadmap 2020/2021

CPU
2.0G+ / 16C
8DDR4 / 128PCIE3
Dual Die / Dual Socket
16nm TSMC


Stand alone GPU
~ 70W
28nm TSMC

source: http://www.gdinsa.org/uploadfile/2020/0326/20200326055237805.pdf

[Tralalak]

HP 268 Pro G1 MT @ ZHAOXIN KaiXian KX-U6780A 2.7GHz


Geekbench 5.1.0 Tryout for Windows x86 (64-bit) (https://browser.geekbench.com/v5/cpu/1876719)
Single-Core Score: 359
Multi-Core Score: 2 391


Geekbench 4.4.1 Tryout for Windows x86 (64-bit (https://browser.geekbench.com/v4/cpu/15421823))
Single-Core Score: 1 829
Multi-Core Score: 8 843

[Tralalak]

Zhaoxin KX-U6780A review: office is stronger than i5-7400
source: https://translate.google.com/translate?hl=sk&sl=zh-CN&tl=en&u=http%3A%2F%2Fnews.mydrivers.com%2F1%2F685%2F685600_all.htm

[Tralalak]

Interesting Japanese article about Centaur CHA x86 processor with AI coprocessor
source: https://translate.google.com/translate?hl=sk&sl=ja&tl=en&u=https%3A%2F%2Fascii.jp%2Felem%2F000%2F004%2F010%2F4010926%2F&sandbox=1

[woodsdog]

Zhaoxin KX-U6780A review: office is stronger than i5-7400
source: https://translate.google.com/translate?hl=sk&sl=zh-CN&tl=en&u=http%3A%2F%2Fnews.mydrivers.com%2F1%2F685%2F685600_all.htm


1. Preface: The latest Zhaoxin KX 6000 processor is a bit stronger than Intel

This is simply a blatant lie proven by their own numbers...
Even compared to a more than 3 Year old i5 7400 (3,5GHz QC) they lose left and right by 30-70%
The SC performance is so absurdly bad (the intel is often 300% faster...), i can't believe there is some kind of Zen Tech in the silicon... Do i confuse this chip with the Zen-Part one? :confused:
Every kind of workload should feel A LOT better on pretty much everything AMD or Intel has to offer. (also: for less Power)

Don't get me wrong, it sure is useable if you do not want to buy US Chips for your Gov Offices... but it sure as hell is not even close to be "stronger than Intel"

[Tralalak]

Why should single-thread performance be important to me for applications like CPU-Z, Cinebench R20, 7-ZIP, wPrime, FritzChess, etc., when I use a Multi-Core processor not single core processor?

Weak single-thread performance only indicates that we are dealing with eight small cores even without hyperthreading as Intel Atom (Pentium J/N, Celeron J/N) cores Airmont, Silvermont, Goldmont or AMD Athlon AM1 (PS4 and XB1) Jaguar, Puma cores.

So it is a "surprise" that even in 2020 the small core has poor performance in single-threaded performance as BIG cores.

Where can I buy a sigle core processor today?

Why wouldn't it be a US chip? After all, the microarchitecture was created by Centaur Technology (https://centtech.com/who-we-are/) (CentaurHauls Family 7 Model 11 Stepping 1) and the iGPU graphics core is IP from S3 Graphics.

There is no Zen technology.

Small core from 65nm to 16nm Centaur Technology
· 65nm CBA VIA Nano 1000/2000
· 65nm CNB VIA Nano 3000
· 40nm CNQ VIA Nano X2 / VIA QuadCore
· 28nm CNR VIA Eden X4 C4xxxx / VIA QuadCore C4xxx
· 28nm CNR = ZX-C C4xxx
· 28nm ZX_D is CNR with reduced pipeline by 5 stages, execution engines were re-balanced, branch prediction unit was reworked and optimized, +2MB L2 cache per four cores (4MB) and removed FSB => new poit-to-point ZPI and DDR4 cotroller with NordBridge PCIe 3.0 + ZX-860 iGPU on same die SoC
· 16nm ZX-E = die shrink ZX_D (ZX-960 iGPU) + SouthBridge I/0 to SoC.

This still the same derivate DNA microarchitecture of small core = Centaur CN ...

vs.

AMD Bulldozer/Piledriver/Excavator = BIG cores
AMD Zen/Zen+/Zen 2 = BIG cores
Intel Haswell, Brodwell, Skylake, Kabylake etc. = BIG cores
Intel Sunny Cove = BIG cores
Centaur CNS (ZX-F) = BIG cores

[woodsdog]

Why should single-thread performance be important to me for applications like CPU-Z, Cinebench R20, 7-ZIP, wPrime, FritzChess, etc., when I use a Multi-Core processor not single core processor?

Weak single-thread performance only indicates that we are dealing with eight small cores even without hyperthreading as Intel Atom (Pentium J/N, Celeron J/N) cores Airmont, Silvermont, Goldmont or AMD Athlon AM1 (PS4 and XB1) Jaguar, Puma cores.

So it is a "surprise" that even in 2020 the small core has poor performance in single-threaded performance as BIG cores.

Where can I buy a sigle core processor today?


defensive much?

Shitty SC Performance in "CPU-Z, Cinebench R20, 7-ZIP, wPrime, FritzChess, etc" means shitty SC Performance everywhere and with everything..

SC performance is a very important aspect in a typical non-Gaming, non-Numbercrunching use of a CPU.
Nobody likes to wait 10 Seconds to start Firefox, Outlook, etc... Nobody likes to wait SECONDS for every Click on a website. In a typical office environment you have lots and lots of small SC Workload. Lots of slow cores are inferior to a few fast ones for this kind of workload.

They said "The latest Zhaoxin KX 6000 processor is a bit stronger than Intel" and benchmark a 2020 8-Core Chip vs a 2017 4-Core Chip and lose with awful bad numbers. Im not sure why they picked this fight but here we are.

There is no Zen technology.
You are right, the "Chinese Zen CPU" from my memorys is not from VIA, its the Hygon Dhyana. I confused this two. That is also the reason i expected so much more from the Via Chip which is "stronger than intel".


Why wouldn't it be a US chip? After all, the microarchitecture was created by Centaur Technology (CentaurHauls Family 7 Model 11 Stepping 1) and the iGPU graphics core is IP from S3 Graphics.
It's not a US Chip anymore because the Development is taking place in China to have silicon where the western backdoors are replaced by chinese ones ;)
The Groundwork did take place on US soil - yes.

And again, i'm just pointing out that their Headline is a lie by their own numbers. The Chips are useable for certain Office workloads but not "fast" or even competitive. These are the facts, there is just nothing to argue.

:love3:

[Tralalak]

An Atom processor (by analogy KX-6000) with a small cores is slower than Core processor (by analogy Centaur CHA) with a BIG cores.

After all, this is basic knowledge. Causality.

Why does it surprise someone?

[woodsdog]

An Atom processor (by analogy KX-6000) with a small cores is slower than Core processor (by analogy Centaur CHA) with a BIG cores.

After all, this is basic knowledge. Causality.
get off your high horse bro, this is a forum to exchange information and opinions, not a lecture on campus.


Why does it surprise someone?
I literally gave you the answer to this question in my last post. Did you not read it? Do you just want to be smug?

[Loeschzwerg]

Wenn alles was du beizutragen hast dein Gesäusel über die Headline ist... Hier ist der passende Thread dafür: https://www.forum-3dcenter.org/vbulletin/showthread.php?t=577870

[robbitop]

Weiß jemand zu wann die angekündigten High Performance Kerne kommen sollen mit ~Zen Performance? :)
Ist der Fertigungsprozess und der geplante Takt schon bekannt? Ggf. ob man diese mit Vorgängercores im Big.Little Stil kombiniert?

[Loeschzwerg]

Also ich würde da erst Ende 2020 / Anfang 2021 mit entsprechenden Produkten rechnen, hält man sich die Entwicklungsgeschichte von KX-5000/KX-6000 vor Augen.
2.5GHz und 16nm sind zumindest laut Roadmap gesetzt, wie auch der grobe Zeitrahmen "2020".

Von VIA sieht man evtl. früher etwas, wenn auch mit Fokus auf Server.

[Tralalak]

k6A696ebqhc

[Loeschzwerg]

My boards have arrived today :)

70110

[rasti666]

Hi

Where did you buy this motherboard? How much did it cost ? I've been looking for something newer than the VIA EPIA-M920 for years - and I can't buy anywhere.

[Tralalak]

My boards have arrived today :)

70110


:up::wave2::massa::weg:

EDIT:

This is your Geekbench 4 and Geekbench 5 results ?

https://browser.geekbench.com/v4/cpu/15462934

https://browser.geekbench.com/v5/cpu/2042186

[Loeschzwerg]

Hi

Where did you buy this motherboard? How much did it cost ? I've been looking for something newer than the VIA EPIA-M920 for years - and I can't buy anywhere.

I imported it from a Taobao seller. Roughly 570€ per board, but shipping, customs and agent also add to the bill.

:up::wave2::massa::weg:

EDIT:

This is your Geekbench 4 and Geekbench 5 results ?

https://browser.geekbench.com/v4/cpu/15462934

https://browser.geekbench.com/v5/cpu/2042186

Nope, those are my results:
https://browser.geekbench.com/v4/cpu/15466403
https://browser.geekbench.com/v5/cpu/2067343
https://valid.x86.fr/pc5ryy

Quick and dirty system setup with some first results. Memory is only running at 2400, the board doesn't use the XMP profile of my GSkill DIMMs. Will see if I can find some modules with a better JEDEC profile.

I didn't expect the IGP to show no performance improvement at all, at least when it comes to 3DMark06. S3 Graphics screen toys report a GPU clock of 650MHz. I'm pretty sure the C-960 does still have 32SP, this would fit the results and GPU clock.

Edit: The system itself feels quite snappy, definitely an improvement compared to a EPIA M920-20Q.

[Benutzername]

Quick and dirty system setup with some first results. Memory is only running at 2400, the board doesn't use the XMP profile of my GSkill DIMMs. Will see if I can find some modules with a better JEDEC profile.

The UEFI is so barebones, that you cannot enter RAM timings by hand?


And wow, spending 570€ for importing those is some real curiosity.

[Screemer]

And wow, spending 570€ for importing those is some real curiosity.
2x :D

[y33H@]

Well, he may not have ordered the boards only for his own fun :biggrin:

[Screemer]

augsburg-berlin-express

[Tralalak]

Zhaoxin KX-U6780A 2.7GHz

https://valid.x86.fr/cache/screenshot/pc5ryy.png"]https://valid.x86.fr/cache/screenshot/pc5ryy.png

CPU-Z 1.92 (Benchmark Version 17.01.64)
Single-Thread: 181 (+14,70%)
Multi-Thread (8T): 1442 (+36,36%)
source: https://valid.x86.fr/pc5ryy


Geekbench 4.4.2 Tryout for Windows x86 (64-bit)
Single-Core Score: 1820 (96,86%)
Multi-Core Score: 9093 (+8,06%)
source: https://browser.geekbench.com/v4/cpu/15466403


Geekbench 5.1.1 Tryout for Windows x86 (64-bit)
Single-Core Score: 362 (93,06%)
Multi-Core Score: 2387 (+8,50%)
source: https://browser.geekbench.com/v5/cpu/2067343


vs


AMD FX-8300 3.3GHz @ 2.7GHz

CPU-Z 1.92 (Benchmark Version 17.01.64)
Single-Thread: 158,1 (87,18%)
Multi-Thread (8T): 1057,2 (73,34%)
https://i.imgur.com/WOz5B9p.jpg


Geekbench 4.4.2 Tryout for Windows x86 (64-bit)
Single-Core Score: 1879 (+3,24%)
Multi-Core Score: 8415 (92,54%)
source: https://imgur.com/5VCNrAv


Geekbench 5.1.1 Tryout for Windows x86 (64-bit)
Single-Core Score: 389 (+7,46%)
Multi-Core Score: 2200 (92,17%)
source: https://imgur.com/wralH10

[Benutzername]

The AMD FX of course makes the KX-U6780A look good, the FX processors are not built for high instructions per cycle. Do you have any comparable intel CPUs for benchmarking avaiable?

[Loeschzwerg]

The UEFI is so barebones, that you cannot enter RAM timings by hand?

Jup, unfortunately that's the case :( Think of a OEM UEFI (Dell, HP, Lenovo, Fujitsu...)


And wow, spending 570€ for importing those is some real curiosity.

:D for science

augsburg-berlin-express

Anfang nächster Woche, ich möchte nur noch die Funktion beider Boards sicherstellen.

[Tralalak]

Today Geekbench updated results and you can view results Zhaoxin KX-U6780A 2.7GHz vs AMD FX-8300 3.3GHz @ 2.7GHz on their web here:

Geekbench 5.1.1 Tryout
https://browser.geekbench.com/v5/cpu/compare/2070351?baseline=2067343


Geekbench 4.4.2 Tryout
https://browser.geekbench.com/v4/cpu/compare/15466947?baseline=15466403

[Loeschzwerg]

@y33H@: Both boards are working :)

@all: Found two DDR4 2666 DIMMs, here are the results.
https://valid.x86.fr/ibtfda
https://browser.geekbench.com/v4/cpu/15469046
https://browser.geekbench.com/v5/cpu/2080204

Memory timings are worse, so the GSKill memory I used yesterday seems to be a better choice.

In the UEFI there is a UMA option for the IGP. Setting it from "normal" to "high" boosts the GPU clock (650MHz -> 1000MHz)...

70139 70140

...resulting in a nice performance boost in 3DMark06 =)

70141 70142

[Loeschzwerg]

Firestrike 650MHz vs 1000MHz:
70143 70144

[y33H@]

Intel HD Graphics 620 is little less than 900 ptk, just for comparison.

[rasti666]

Loeschzwerg - thank you for the information. I plan to buy this motherboard ;-). I have one more question. How did you connect the power supply? Because I don't understand it at all. I don't see 24-pin input.

[Tralalak]

Tro4Rco8s3w

[Benutzername]

Loeschzwerg - thank you for the information. I plan to buy this motherboard ;-). I have one more question. How did you connect the power supply? Because I don't understand it at all. I don't see 24-pin input.

Looks like Loeschzwerg connected only the 4-pin ATX12V plug in his photograph.

[rasti666]

I saw it too. The ATX power supply should not start the computer in this situation. Loeschzwerg had to do something else. I don't know what he did, that's why I asked ;-). How exactly is the power supply connected ?

[Loeschzwerg]

Step by step:
1) Use a power supply with an On/Off switch on the back (makes things easier and safer)
2) Like Benutzername already mentioned, connect the 4-Pin ATX12V (on modern power supply it's a 4+4 connector) to the motherboard
3) Short PS-ON pin (green wire) on the ATX main connector to ground (a black wire). You can use a paper clip (that's what I did for my quick&dirty test) or wire a switch to those pins... There are many options.
4) Turn on the power supply (On/Off switch on the back)
5) Power on the motherboard by pressing the small button on the PCB

[Tralalak]

Hewlett Packard stellt neuen Komplett-PC mit China-CPU vor
Quelle: https://www.pcgameshardware.de/CPU-CPU-154106/News/chinesischer-Prozessor-neuer-Komplett-PC-Hewlett-Packard-1349804/


HP 268 Pro G1: Komplett-PC mit chinesischem Achtkern-Prozessor
Quelle: https://www.heise.de/newsticker/meldung/HP-268-Pro-G1-Komplett-PC-mit-chinesischem-Achtkern-Prozessor-4717638.html


HP 268 Pro G1: HP verkauft Client-PC mit chinesischer KaiXian-CPU
Quelle: https://www.computerbase.de/2020-05/hp-268-pro-g1-client-pc-zhaoxin-kaixian-kx-u6780a/

[y33H@]

Well, that's old news ^^

[Tralalak]

Well, that's old news ^^
Yes, for this forum is old news.
For other websites in Europe is current news.

[Benutzername]

Yes, for this forum is old news.
For other websites in Europe is current news.


Interesting that the news makes the rounds. In recent months several sites, youtube channels have tested ZhaoXhin boards. So it really spreads that there is something new from China (with Centaur's and VIA's help)

----------------------------
Step by step:
1) Use a power supply with an On/Off switch on the back (makes things easier and safer)
2) Like Benutzername already mentioned, connect the 4-Pin ATX12V (on modern power supply it's a 4+4 connector) to the motherboard
3) Short PS-ON pin (green wire) on the ATX main connector to ground (a black wire). You can use a paper clip (that's what I did for my quick&dirty test) or wire a switch to those pins... There are many options.
4) Turn on the power supply (On/Off switch on the back)
5) Power on the motherboard by pressing the small button on the PCB


As the board only uses 12V any 12V power supply with enough Watts should also work with the appropiate plug fitted me thinks.

[Loeschzwerg]

A DC-DC power supply with barrel plug comes with the board and it's enough to drive the system with a NVME SSD, SATA, Wifi card and so on. Unfortunately the needed power cables from motherboard to SATA-power/Molex 4-pin weren't included, so I went for the ATX power supply solution. I had planned to add a dGPU (6-pin PCIe needed) anyway, so no big deal for me :)

Currently I'm installing Windows 10 on a Samsung PM961 256GB NVME SSD.

---
Edit:
MrH OpenCL Benchmark
70148

CrystalDiskMark 7.0.0 (Samsung PM961 256GB)
70149

[Tralalak]

Rumours said that new Zhaoxin Stand alone GPU ~70W 28nm TSMC have performace around ~GTX1050 and two those graphics card around ~GTX1070.
The graphics card will be important for a ZX-F KH-40000 processor which does not have a integrated graphics card in to SoC.

http://www.gdinsa.org/uploadfile/2020/0326/20200326055237805.pdf

[Loeschzwerg]

Rumours said that new Zhaoxin Stand alone GPU ~70W 28nm TSMC have performace around ~GTX1050 and two those graphics card around ~GTX1070.
The graphics card will be important for a ZX-F KH-40000 processor which does not have a integrated graphics card in to SoC.

http://www.gdinsa.org/uploadfile/2020/0326/20200326055237805.pdf

On page 13 (roadmap) there is also a "hidden" information on the top left, with font color white on a white background :freak:

2.0G+/4C
DX12/15W
16nm TSMC

So they are also planning a small, low power SoC with a DX12 capable IGP.

[rasti666]

Loeschzwerg - thank you very much for the information about the power supply. They will be very useful.

[Tralalak]

On page 13 (roadmap) there is also a "hidden" information on the top left, with font color white on a white background :freak:

2.0G+/4C
DX12/15W
16nm TSMC

So they are also planning a small, low power SoC with a DX12 capable IGP.

Yes because notebooks also ultra-thin notebooks on the way:

https://m.mydrivers.com/newsview/688128.html
http://www.icsmart.cn/36771/

P. S. HOW IS HWINFO64 DETECTION?

[y33H@]

Yeah, there was a Lenovo a few months ago: https://imgur.com/a/kIK8bmw

[Tralalak]

Yeah, there was a Lenovo a few months ago: https://imgur.com/a/kIK8bmw
Well, that's old news ^^

But anyway.
It's a shame that they have only 16nm FCC. And that's not the best 16nm from TSMC.

[y33H@]

It's cheaper ^^

[Benutzername]

It's cheaper ^^

Exactly. I doubt they make much profit with their processors at the moment so contract for a cheap process with wafers available at TSMC.


Ona all important question has not been asked yet:

Does it run Crysis?



:biggrin:

[Tralalak]

It's cheaper ^^
Yes. But then they can't reach high frequencies.

Glenn henry said here 17:24
https://www.youtube.com/watch?v=5Z7cmyYakAw

[y33H@]

I think it's a reasonable trade-off.

[y33H@]

Ona all important question has not been asked yet:

Does it run Crysis?

:biggrin:Yes, even in software mode :freak:

[Tralalak]

In the past (2011), I tried the Crysis 2 on the VIA Nano X2 1.6GHz and Nvidia GeForce GT430 1GB DDR3. I didn't have a powerful graphics card then.

5e9xZfOBHVA

[Loeschzwerg]

P. S. HOW IS HWINFO64 DETECTION?

Not that great
70161

---

CB R11.5 and R20 results
70162 70163

[Tralalak]

Not that great
70161


You wrote to Martin M. from HWiNFO? If no pls send him (mm@hwinfo.com) debug file with debug mode and prnt scrn. He support you.

[rasti666]

I see that RAM clock speed is always 2400 MHz. Is this the maximum possibility? CPU-Z and HWiNFO show the same.

Similarly, the G.Skill F4-2800C18-8GRS (DDR4 2800 MHz) tests show - start only at 2400 MHz.

Are the maximum RAM speeds on the motherboard ?

[Loeschzwerg]

Current max. of the board is DDR4-2666. You need RAM with JEDEC profiles. Since my G.Skill is a XMP memory and the highest JEDEC profile is only 2400, I'm stuck with 2400.
I also have some regular Samsung DDR4-2666, but the timings are worse which does have a negative impact on performance.

My CPU-Z validation with 2666 memory in use:
https://valid.x86.fr/ibtfda

Edit:
You wrote to Martin M. from HWiNFO? If no pls send him (mm@hwinfo.com) debug file with debug mode and prnt scrn. He support you.

Done ;)

[Loeschzwerg]

AIDA64 GPGPU and Cache&Memory results. Also compared ZX C960 running at 650MHz vs 1000MHz.

70193 70194 70195

[Tralalak]

Relative performance of Zhaoxin's processors in SPECint_2006


Single Thread: ZX-A (1) vs ZX-C (1,4x) vs KX-5000 (1,8x) vs KX-6000 (2,7x) vs KX-7000 (4,4x)

Multi Thread: ZX-A (1) vs ZX-C (2,3x) vs KX-5000 (5,8x) vs KX-6000 (8,1x) vs KX-7000 (31x)

source: https://read01.com/J8j2MyQ.html#.Xr7Eu8DgqUl


A large increase in the performance of the new microarchitecture.
ZX-A / ZX-C / KX-5000 / KX-6000 = small cores
KX-7000 = BIG cores

[Loeschzwerg]

Inter core latency of the KX-U6780A:
U0-U1 Latenzzeit Daten : 54.0ns
U0-U2 Latenzzeit Daten : 54.3ns
U0-U3 Latenzzeit Daten : 54.0ns
U0-U4 Latenzzeit Daten : 153.0ns
U0-U5 Latenzzeit Daten : 152.4ns
U0-U6 Latenzzeit Daten : 148.9ns
U0-U7 Latenzzeit Daten : 146.9ns
U1-U2 Latenzzeit Daten : 53.7ns
U1-U3 Latenzzeit Daten : 53.9ns
U1-U4 Latenzzeit Daten : 146.6ns
U1-U5 Latenzzeit Daten : 147.4ns
U1-U6 Latenzzeit Daten : 147.6ns
U1-U7 Latenzzeit Daten : 146.4ns
U2-U3 Latenzzeit Daten : 53.8ns
U2-U4 Latenzzeit Daten : 147.1ns
U2-U5 Latenzzeit Daten : 148.1ns
U2-U6 Latenzzeit Daten : 147.4ns
U2-U7 Latenzzeit Daten : 146.7ns
U3-U4 Latenzzeit Daten : 149.2ns
U3-U5 Latenzzeit Daten : 148.9ns
U3-U6 Latenzzeit Daten : 147.8ns
U3-U7 Latenzzeit Daten : 148.8ns
U4-U5 Latenzzeit Daten : 53.8ns
U4-U6 Latenzzeit Daten : 53.8ns
U4-U7 Latenzzeit Daten : 53.8ns
U5-U6 Latenzzeit Daten : 53.7ns
U5-U7 Latenzzeit Daten : 53.8ns
U6-U7 Latenzzeit Daten : 53.7ns

Complete result of the "Multi-Core Efficiency" test


SiSoftware Sandra

Benchmarkresultate
Bandbreite Kern-Kern : 12.64GB/s
Ergebnisse : Höhere Werte sind besser.
Base 2 Ergebnis Multiplikatoren : 1GB(/s) = 1024MB(/s), 1MB(/s) = 1024kB(/s), 1kB(/s) = 1024 bytes(/s), usw.

Benchmarkresultate
Latenzzeit Kern-Kern : 107.8ns
Ergebnisse : Geringere Werte sind besser.
Base 10 Ergebnis Multiplikatoren : 1s = 1000ms, 1ms = 1000µs, 1µs = 1000ns, usw.

Leistung pro Thread
Bandbreite Kern-Kern : 1.58GB/s
Threadanzahl : 8
Ergebnisse : Höhere Werte sind besser.
Base 2 Ergebnis Multiplikatoren : 1GB(/s) = 1024MB(/s), 1MB(/s) = 1024kB(/s), 1kB(/s) = 1024 bytes(/s), usw.

Gegenüberstellung Leistung und Geschwindigkeit
Bandbreite Kern-Kern : 4.80MB/s/MHz
Ergebnisse : Höhere Werte sind besser.
Latenzzeit Kern-Kern : 0.40ns/MHz
Ergebnisse : Geringere Werte sind besser.

Benchmarkabbruch
Prozessorähnlichkeit : U0-U1 U2-U3 U4-U5 U6-U7
U0-U1 Latenzzeit Daten : 54.0ns
U0-U2 Latenzzeit Daten : 54.3ns
U0-U3 Latenzzeit Daten : 54.0ns
U0-U4 Latenzzeit Daten : 153.0ns
U0-U5 Latenzzeit Daten : 152.4ns
U0-U6 Latenzzeit Daten : 148.9ns
U0-U7 Latenzzeit Daten : 146.9ns
U1-U2 Latenzzeit Daten : 53.7ns
U1-U3 Latenzzeit Daten : 53.9ns
U1-U4 Latenzzeit Daten : 146.6ns
U1-U5 Latenzzeit Daten : 147.4ns
U1-U6 Latenzzeit Daten : 147.6ns
U1-U7 Latenzzeit Daten : 146.4ns
U2-U3 Latenzzeit Daten : 53.8ns
U2-U4 Latenzzeit Daten : 147.1ns
U2-U5 Latenzzeit Daten : 148.1ns
U2-U6 Latenzzeit Daten : 147.4ns
U2-U7 Latenzzeit Daten : 146.7ns
U3-U4 Latenzzeit Daten : 149.2ns
U3-U5 Latenzzeit Daten : 148.9ns
U3-U6 Latenzzeit Daten : 147.8ns
U3-U7 Latenzzeit Daten : 148.8ns
U4-U5 Latenzzeit Daten : 53.8ns
U4-U6 Latenzzeit Daten : 53.8ns
U4-U7 Latenzzeit Daten : 53.8ns
U5-U6 Latenzzeit Daten : 53.7ns
U5-U7 Latenzzeit Daten : 53.8ns
U6-U7 Latenzzeit Daten : 53.7ns
Inter-Core Bandbreite @ 1x 64bytes : 1.15GB/s
Inter-Core Bandbreite @ 4x 64bytes : 1.79GB/s
Inter-Core Bandbreite @ 4x 256bytes : 6.34GB/s
Inter-Core Bandbreite @ 4x 1024bytes : 18GB/s
Inter-Core Bandbreite @ 4x 4kB : 29.58GB/s
Inter-Core Bandbreite @ 16x 4kB : 34.35GB/s
Inter-Core Bandbreite @ 4x 64kB : 50.24GB/s
Inter-Core Bandbreite @ 16x 64kB : 50.27GB/s
Inter-Core Bandbreite @ 8x 256kB : 46.86GB/s
Inter-Core Bandbreite @ 4x 1024kB : 11GB/s
Inter-Core Bandbreite @ 8x 1024kB : 8.48GB/s
Inter-Core Bandbreite @ 8x 4MB : 6.31GB/s

Leistungsteststatus
Ergebnis-ID : ZHAOXIN KaiXian KX-U6780A@2.7GHz (8C 2.7GHz, 2x 4MB L2)
Microcode
System : Zhaoxin ZXE CRB ZXE (Zhaoxin HX002ED0)
Kompatibel mit Plattform : x64
Verwendete Pufferung : Nein
Threadanzahl : 8
Systemtakt : 14.32MHz
Speicherseitengröße : 2MB

Prozessor
Modell : ZHAOXIN KaiXian KX-U6780A@2.7GHz
Geschwindigkeit : 2.7GHz
Minimale/Maximale/Turbo Geschwindigkeit : k.A. - 2.7GHz
Kerne pro Prozessor : 8 Einheit(en)
Threads pro Kern : 1 Einheit(en)
Front Side Bus Geschwindigkeit : 100MHz
Revision/Stepping : 3B / 1
Interner (L1D) Datencache : 8x 32kB, 8-Weg, 64bytes Zeilengröße
Interner (L1I) Anweisungscache : 8x 32kB, 8-Weg, 64bytes Zeilengröße
L2D-Datencache/Einheitscache : 2x 4MB, 16-Weg, Voll integriert, 64bytes Zeilengröße, 4 Thread(s)

[y33H@]

Looks like that then :biggrin:

70281

[Benutzername]

Looks like that then :biggrin:

70281

So basically two clusters of four cores and the latency shoots up when you have to talk between the two clusters?

[y33H@]

Correct - and as expected.

[Poulton]

ZX-A / ZX-C / KX-5000 / KX-6000 = small cores
KX-7000 = BIG cores
So the KX-U6780A is basicly the counterpart to the Intel Atom (even if they name the Desktop-Version not so anymore)?


My CPU-Z validation with 2666 memory in use:
https://valid.x86.fr/ibtfda
Damit liegt er bei dem Test auf dem Niveau eines Apollo Lake Pentium J4205 hatte, den ich 2017 ein 3/4 Jahr verwendete.
https://valid.x86.fr/q1wgta

70306

Noch ein paar Tests die ich mit dem damals gemacht hatte:

7zip Single: Gesamt: 2067 MIPS ; Kompr.: 1767 MIPS ; Dekompr.: 2367 MIPS
https://abload.de/img/7zipsingleuiy64.jpg

7zip Multi: Gesamt: 7582 MIPS ; Kompr.: 6405 MIPS ; Dekompr.: 8760 MIPS
https://abload.de/img/7zipmultileayf.jpg

3DMark 2013:
Ice Storm: Score: 37005 ; Graphics: 43512 ; Physics: 24292
http://www.3dmark.com/3dm/20061548

Cloud Gate: Score: 4143 ; Graphics: 5457 ; Physics: 2249
https://www.3dmark.com/3dm/20061473

Sky Diver: Score: 2474 ; Graphics: 2367 ; Physics: 3306 ; Combined: 2391
http://www.3dmark.com/3dm/20061078

Fire Strike: Score: 643 ; Graphics: 691 ; Physics: 3173 ; Combined: 238
http://www.3dmark.com/3dm/20061308

Time Spy: Score: 156 ; Graphics: 137 ; CPU: 828
http://www.3dmark.com/3dm/20060446

HWBOT x265 v1.1.1: 3,93fps
https://abload.de/img/hwbotx265v1.1.14cl3o.jpg

VeraCrypt 50MB: https://abload.de/img/veracrypt50mbkwzrz.jpg
VeraCrypt 1GB: https://abload.de/img/veracrypt1gbyal30.jpg

POV-Ray v3.7 Multi: 511,96 PPS
https://abload.de/img/pov-ray3.7multik3agx.jpg

POV-Ray v3.7 Single: 144,82 PPS
https://abload.de/img/pov-ray3.7singlevgyoi.jpg

Cinebench 11.5: Single: 0,69 Punkte ; Multi: 2,50 Punkte
https://abload.de/img/cinebench11.50yb3a.jpg

AIDA64 Extreme 5.90.4200: https://abload.de/img/cachemem5.90.4200i3lna.png

HWiNFO64: https://abload.de/img/hwinfo64vxb59.jpg

Leider hatte ich die Tests nicht wiederholt, als später eine GT1030 dazukam. Einzig für Half Life 2 Lost Coast hätte ich da einen Vergleich:

Grafikeinstellungen: https://abload.de/image.php?img=unbenannt-15suaz.png
Befehle die vorher in die Konsole gekommen sind: „disconnect“ und „fps_max 10000“

IGP 1280x720: 71.62
https://abload.de/img/20170228152519_1crsch.jpg

IPG 1920x1080: 42.16
https://abload.de/img/20170228153904_13bsss.jpg

GT1030 1280x720: 160.38
https://abload.de/img/20170913182711_1uzbda.jpg

GT1030 1920x1080: 155.39
https://abload.de/img/20170913183533_13daho.jpg


Schade dass das Asrock J5040 (Gemini Lake Refresh) nicht verfügbar ist oder es gar etwas mit dem Nachfolger auf Tremont-Basis gibt. AMD hat ja leider den Puma-Käfig geschlossen. Bliebe noch Nvidia, die sich bei ARM austoben (https://www.phoronix.com/scan.php?page=article&item=nvidia-xavier-carmel&num=1).

[y33H@]

Ich hab beim Test des KX-U6780A nen Pentium J5005 (Gemini Lake Atom) mit drin.

[Loeschzwerg]

For comparison the core latency of the VIA CNR Quadcore C4650 (2GHz):

U0-U1 Latenzzeit Daten : 70.1ns
U0-U2 Latenzzeit Daten : 72.6ns
U0-U3 Latenzzeit Daten : 71.2ns
U1-U2 Latenzzeit Daten : 69.9ns
U1-U3 Latenzzeit Daten : 70.6ns
U2-U3 Latenzzeit Daten : 69.9ns



SiSoftware Sandra

Benchmarkresultate
Bandbreite Kern-Kern : 3.16GB/s
Ergebnisse : Höhere Werte sind besser.
Base 2 Ergebnis Multiplikatoren : 1GB(/s) = 1024MB(/s), 1MB(/s) = 1024kB(/s), 1kB(/s) = 1024 bytes(/s), usw.

Benchmarkresultate
Latenzzeit Kern-Kern : 70.7ns
Ergebnisse : Geringere Werte sind besser.
Base 10 Ergebnis Multiplikatoren : 1s = 1000ms, 1ms = 1000µs, 1µs = 1000ns, usw.

Leistung pro Thread
Bandbreite Kern-Kern : 809MB/s
Threadanzahl : 4
Ergebnisse : Höhere Werte sind besser.
Base 2 Ergebnis Multiplikatoren : 1GB(/s) = 1024MB(/s), 1MB(/s) = 1024kB(/s), 1kB(/s) = 1024 bytes(/s), usw.

Gegenüberstellung Leistung und Geschwindigkeit
Bandbreite Kern-Kern : 1.62MB/s/MHz
Ergebnisse : Höhere Werte sind besser.
Latenzzeit Kern-Kern : 0.35ns/MHz
Ergebnisse : Geringere Werte sind besser.

Benchmarkabbruch
Prozessorähnlichkeit : U0-U1 U2-U3
U0-U1 Latenzzeit Daten : 70.1ns
U0-U2 Latenzzeit Daten : 72.6ns
U0-U3 Latenzzeit Daten : 71.2ns
U1-U2 Latenzzeit Daten : 69.9ns
U1-U3 Latenzzeit Daten : 70.6ns
U2-U3 Latenzzeit Daten : 69.9ns
Inter-Core Bandbreite @ 1x 64bytes : 430MB/s
Inter-Core Bandbreite @ 4x 64bytes : 702MB/s
Inter-Core Bandbreite @ 4x 256bytes : 2.47GB/s
Inter-Core Bandbreite @ 4x 1024bytes : 6.68GB/s
Inter-Core Bandbreite @ 4x 4kB : 11.15GB/s
Inter-Core Bandbreite @ 16x 4kB : 12.19GB/s
Inter-Core Bandbreite @ 4x 64kB : 17.91GB/s
Inter-Core Bandbreite @ 16x 64kB : 14.23GB/s
Inter-Core Bandbreite @ 8x 256kB : 2.17GB/s
Inter-Core Bandbreite @ 4x 1024kB : 1.43GB/s
Inter-Core Bandbreite @ 8x 1024kB : 1.43GB/s
Inter-Core Bandbreite @ 8x 4MB : 1.37GB/s

Leistungsteststatus
Ergebnis-ID : VIA QuadCore C4650@2.0GHz (4C 2GHz, 2MB L2)
Microcode
System : VIA EPIA-M920
Kompatibel mit Plattform : x64
Verwendete Pufferung : Nein
Threadanzahl : 4
Systemtakt : 14.32MHz
Speicherseitengröße : 2MB

Prozessor
Modell : VIA QuadCore C4650@2.0GHz
Geschwindigkeit : 2GHz
Minimale/Maximale/Turbo Geschwindigkeit : 1GHz - 2GHz
Kerne pro Prozessor : 4 Einheit(en)
Threads pro Kern : 1 Einheit(en)
Front Side Bus Geschwindigkeit : 267MHz
Typ : Laptop/Netbook
Revision/Stepping : F / E
Interner (L1D) Datencache : 4x 32kB, 8-Weg, 64bytes Zeilengröße
Interner (L1I) Anweisungscache : 4x 32kB, 8-Weg, 64bytes Zeilengröße
L2D-Datencache/Einheitscache : 2MB, 16-Weg, 64bytes Zeilengröße, 4 Thread(s)


------
Edit:
..and Eden X4 C4250 (1.2GHz up to 1.73GHz; VIA Artigo A1250)

U0-U1 Latenzzeit Daten : 107.6ns
U0-U2 Latenzzeit Daten : 107.6ns
U0-U3 Latenzzeit Daten : 107.5ns
U1-U2 Latenzzeit Daten : 97.1ns
U1-U3 Latenzzeit Daten : 98.3ns
U2-U3 Latenzzeit Daten : 97.5ns



SiSoftware Sandra

Benchmarkresultate
Bandbreite Kern-Kern : 2.21GB/s
Ergebnisse : Höhere Werte sind besser.
Base 2 Ergebnis Multiplikatoren : 1GB(/s) = 1024MB(/s), 1MB(/s) = 1024kB(/s), 1kB(/s) = 1024 bytes(/s), usw.

Benchmarkresultate
Latenzzeit Kern-Kern : 102.6ns
Ergebnisse : Geringere Werte sind besser.
Base 10 Ergebnis Multiplikatoren : 1s = 1000ms, 1ms = 1000µs, 1µs = 1000ns, usw.

Leistung pro Thread
Bandbreite Kern-Kern : 565MB/s
Threadanzahl : 4
Ergebnisse : Höhere Werte sind besser.
Base 2 Ergebnis Multiplikatoren : 1GB(/s) = 1024MB(/s), 1MB(/s) = 1024kB(/s), 1kB(/s) = 1024 bytes(/s), usw.

Gegenüberstellung Leistung und Geschwindigkeit
Bandbreite Kern-Kern : 1.54MB/s/MHz
Ergebnisse : Höhere Werte sind besser.
Latenzzeit Kern-Kern : 0.70ns/MHz
Ergebnisse : Geringere Werte sind besser.

Benchmarkabbruch
Prozessorähnlichkeit : U0-U3 U1-U2
U0-U1 Latenzzeit Daten : 107.6ns
U0-U2 Latenzzeit Daten : 107.6ns
U0-U3 Latenzzeit Daten : 107.5ns
U1-U2 Latenzzeit Daten : 97.1ns
U1-U3 Latenzzeit Daten : 98.3ns
U2-U3 Latenzzeit Daten : 97.5ns
Inter-Core Bandbreite @ 1x 64bytes : 257MB/s
Inter-Core Bandbreite @ 4x 64bytes : 415MB/s
Inter-Core Bandbreite @ 4x 256bytes : 1.47GB/s
Inter-Core Bandbreite @ 4x 1024bytes : 4GB/s
Inter-Core Bandbreite @ 4x 4kB : 6.7GB/s
Inter-Core Bandbreite @ 16x 4kB : 7.3GB/s
Inter-Core Bandbreite @ 4x 64kB : 10.58GB/s
Inter-Core Bandbreite @ 16x 64kB : 9.29GB/s
Inter-Core Bandbreite @ 8x 256kB : 2.12GB/s
Inter-Core Bandbreite @ 4x 1024kB : 1.32GB/s
Inter-Core Bandbreite @ 8x 1024kB : 1.32GB/s
Inter-Core Bandbreite @ 8x 4MB : 1.25GB/s

Leistungsteststatus
Ergebnis-ID : VIA Eden X4 C4250@1.2+GHz (4C 1.47GHz/1.73GHz, 2MB L2)
Microcode
System : VIA Embedded ARTiGO A1250
Kompatibel mit Plattform : x64
Verwendete Pufferung : Nein
Threadanzahl : 4
Systemtakt : 14.32MHz
Speicherseitengröße : 2MB

Prozessor
Modell : VIA Eden X4 C4250@1.2+GHz
Geschwindigkeit : 1.47GHz
Minimale/Maximale/Turbo Geschwindigkeit : 1GHz - 1.73GHz
Kerne pro Prozessor : 4 Einheit(en)
Threads pro Kern : 1 Einheit(en)
Front Side Bus Geschwindigkeit : 267MHz
Typ : Laptop/Netbook
Revision/Stepping : F / E
Interner (L1D) Datencache : 4x 32kB, 8-Weg, 64bytes Zeilengröße
Interner (L1I) Anweisungscache : 4x 32kB, 8-Weg, 64bytes Zeilengröße
L2D-Datencache/Einheitscache : 2MB, 16-Weg, 64bytes Zeilengröße, 4 Thread(s)

[y33H@]

Teaser!

70331 70332 70336

70334 70333

THX @ Loeschzwerg =)

[Tralalak]

Cinebench R20 (Multi Core)

Loeschzwerg: 933 cb (+5,30%)
y33H@: 886 cb


The difference of 47 cb is large for the same boards. Maybe Loeschzwerg had better memory modules than y33H ?

[y33H@]

I was using DDR4-2666 @ 18-18-18-43, he seems to utilize DDR4-2400 @ 18-15-17-39.

[Poulton]

Would it be possible to run Cinebench 20 Multithread with only 4 Cores, for a better comparision with the J5005?
70341

[y33H@]

Sure, but simply divide by two and you should be at least close :freak:

EDIT

70342

[Tralalak]

Hi guys. Could you also run someone benchmark in Sisoft Sandra?
We can see then results in ranker: https://ranker.sisoftware.co.uk/top_run_all.php?q=d6ebd3&l=en

Processor Arithmetic
Processor Multi-Media
Processor Cryptography (Standard Security)
Processor Multi-Core Efficiency

Memory Bandwidth
Data Cache Bandwidth
Video Shader Compute

(Optional)
Processor Financial Analysis (Normal/Single Precision)
Processor Financial Analysis (High/Double Precision)
Processor Scientific Analysis (Normal/Single Precision)
Processor Scientific Analysis (High/Double Precision)
Processor Image Processing (Normal/Single Precision)
Processor Neural Networks (Normal/Single Precision)
Processor Neural Networks (High/Double Precision)

Thanks

[Poulton]

Would it also be possible to run some Linux-Benchmarks somewhere in the future, like the one that Phoronix did (https://www.phoronix.com/scan.php?page=article&item=nvidia-xavier-carmel&num=1) with the ARM-based Nvidia Jetson AGX Xavier Carmel (8C/8T, 32GB DDR4) against some Intel CPU?

[y33H@]

We did test Linux but no dedicated benchmarks - it simply runs, at least the CPU and software rasterizer (LLVMpipe) for the iGPU. But yes, we may do some benchmarks.

[Tralalak]

2020 ACM/IEEE 47th Annual International Symposium on Computer Architecture (ISCA)
High-Performance Deep-Learning Coprocessor Integrated into x86 SoC with Server-Class CPUs
source: https://www.iscaconf.org/isca2020/papers/466100a015.pdf

[Loeschzwerg]

Zhaoxin KX-U6780A im Test (Golem):
https://www.golem.de/news/zhaoxin-kx-u6780a-im-test-das-kann-chinas-x86-prozessor-2005-148364.html

I've contributed the C4650 results.

[y33H@]

THX @ Loeschzwerg =)

Zhaoxin KX-U6780A review:

https://www.golem.de/news/zhaoxin-kx-u6780a-im-test-das-kann-chinas-x86-prozessor-2005-148364.html

[konkretor]

geiler test!

[Darkstar]

geiler test!Dem schließe ich mich vollumfänglich an! Danke! ;D

[Tralalak]

THX @ Loeschzwerg =)

Zhaoxin KX-U6780A review:

https://www.golem.de/news/zhaoxin-kx-u6780a-im-test-das-kann-chinas-x86-prozessor-2005-148364.html

Atom und Bulldozer schlagen den Megacore
"Einzig bei 7-Zip, vor allem beim Dekomprimieren, kann sich der KX-U6780A pro Takt vom Pentium J5005 absetzen."
And here I stopped reading it...

:hammer:


And I knew exactly what was going to happen. All the editors started to take over. Many people just read the headlines and then the discussions still look like that.

[y33H@]

Ohne diesen Thread und viele seiner Nutzer wäre dieser Test so nicht möglich gewesen, ich gebe den Dank also zurück.

@ Tralalak

What's your problem with the description of the performance? I'm just giving context.

[Igzorn]

Großes Lob an den Verfasser des Artikel.
Wenn man bedenkt wie viele Kerne die CPU hast ist sie zwar wirklich langsam, aber nur um mal mit der Technik zu Spielen hätte ich schon gern so ein Board.
Wenn es den wenigstens Preislich passen würde:redface:

[Tralalak]

Zhaoxin KiaXian KX-U6780A 2.7GHz in the Anandtech CPU 2019 Benchmarks Database
source: https://www.anandtech.com/bench/CPU-2019/2224

[y33H@]

Dualchannel doesn't seem to push the performance, mostly identical to my/Loeschzwerg's data, as expected.

[Tralalak]

Dualchannel doesn't seem to push the performance, mostly identical to my/Loeschzwerg's data, as expected.


Cinebench R20 (Multi Core)

Loeschzwerg: 933 cb (+5,30%)
y33H@: 886 cb (94,96 %)


The difference of 47 cb is large for the same boards. Maybe Loeschzwerg had better memory modules than y33H ?



Geekbench 5.1.1 - Multi-Core Score

Loeschzwerg: Zhaoxin KX-U6780A 2.7GHz (8C/8T) 2387,00 (+3,51 %)
y33H@: Zhaoxin KX-U6780A 2.7GHz (8C/8T) 2306,00 (96,61 %)


Geekbench 5.1.1 Multi-Core Score - Crypto Score

Loeschzwerg: Zhaoxin KX-U6780A 2.7GHz (8C/8T) 3428,00 (+10,12 %)
y33H@: Zhaoxin KX-U6780A 2.7GHz (8C/8T) 3113,00 (90,81 %)


Geekbench 5.1.1 Multi-Core Score - Integer Score

Loeschzwerg: Zhaoxin KX-U6780A 2.7GHz (8C/8T) 2455,00 (+4,82 %)
y33H@: Zhaoxin KX-U6780A 2.7GHz (8C/8T) 2342,00 (95,40 %)


Geekbench 5.1.1 Multi-Core Score - Floating Point Score

Loeschzwerg: Zhaoxin KX-U6780A 2.7GHz (8C/8T) 2138,00 (+2,15 %)
y33H@: Zhaoxin KX-U6780A 2.7GHz (8C/8T) 2093,00 (97,90 %)



I was using DDR4-2666 @ 18-18-18-43, he seems to utilize DDR4-2400 @ 18-15-17-39.




The question is. What would be the results with DDR4-2400 @ 18-15-17-39 in dual channel connection ?




PLUS
My experience is that many benchmarks increase overall performance in dual-channel connections.



Here is a brief example of my laptop. However, other hardware.


y-cruncher - A Multi-Threaded Pi-Program


Intel Core i5 1035G1 and 1x 8GB single-channel SO-DIMM DDR4


https://i.imgur.com/xYGpFjL.jpg


vs


Intel Core i5-1035G1 @ 1.00GHz and 16GB (2x 8GB) dual-channel SO-DIMM DDR4


https://i.imgur.com/xkVCSqI.jpg

[y33H@]

A 1035G is way faster (per core/thread) than the U6780A so it also utilizes more bandwidth. Shorter latencies will help the U6780A but you cannot set them with the Cjoyin board and Ian's dev board is also kinda buggy.

[Tralalak]

zqBSJSAmCHQ

[Benutzername]

https://youtu.be/zqBSJSAmCHQ

Good to hear that he had no real problems getting the board to run and that it is stable under Win10.

at about 12:15 he fiddles around with an intel stock cooler. Do the holes fit an intel type mounitng? The cooler deiivered with the board looks more at home in a tight industrial PC case and even the stock intel cooler should be better. More mass, more fins. Not to forget aftermarket coolers.

[Loeschzwerg]

Do the holes fit an intel type mounitng?

Yes, with the V20 Version of the C1888 board you can use common Intel heatsink designs.

[scx]

THX @ Loeschzwerg =)

Zhaoxin KX-U6780A review:

https://www.golem.de/news/zhaoxin-kx-u6780a-im-test-das-kann-chinas-x86-prozessor-2005-148364.html

Great work! Unfortunately, I found some inaccuracies.


Die vier Kerne sind nativ, wie bei Via auch - zuvor gab es zwei Chips auf einem Träger.


The four cores are native, as with Via - previously there were two chips on a carrier.

This was true for VIA CNQ (2011), because it was basically two Nano X2 in a multi-chip module. However, CNR (2015, samples available in 2014) provides 4 cores on a single die area.

See also:
https://www.forum-3dcenter.org/vbulletin/showthread.php?p=12148644#post12148644


Mit den KX-5000 (ZX-C+) koppelte Zhaoxin dann zwei Quadcore-Dies, um so acht CPU-Kerne zu erhalten.


Zhaoxin then coupled two quad-core dies with the KX-5000 (ZX-C+) in order to obtain eight CPU cores.

First of all, KX-5000 is ZX-D, not ZX-C+. There are 4-cores ZX-C+ CPUs (e.g. C4701 - 4 cores on a single die), as well as 8-cores (e.g. FC-1081 - basically two quad-core CPUs in a multi-chip module).

See also:
http://en.zhaoxin.com/ZXC.aspx?seriesid=5
http://en.zhaoxin.com/ZXC.aspx?seriesid=13


Außerdem wurde die Zhanjiang-Technik um SM3 und SM4 erweitert: Diese Verschlüsselungsverfahren sind erforderlich, damit die Prozessoren in staatlichen Institutionen in China eingesetzt werden dürfen.


In addition, the Zhanjiang technology was expanded to include SM3 and SM4: These encryption methods are required so that the processors can be used in state institutions in China.

This is not false, but it could be better explained: ZX-C is basically CNR, while ZX-C is CNR with additional instruction sets: SM3 and SM4. Both ZX-C and ZX-C+ belongs to the Zhangjiang microarchitecture, which is based on Isaiah II (the same CPUID). In other words, CNR (Isaiah II), ZX-C (Zhangjiang) and ZX-C+ (Zhangjiang) have the same CPUID, although the last one provides support for additional instruction sets: SM3 and SM4.


Waren bisher alle CPUs auf einen I/O-Chip namens ZX-100S


Until now, all CPUs had to rely on an I/O chip called ZX-100S

ZX-A, ZX-B, ZX-C, and ZX-C+ can also work with the VIA VX11(PH) chipset. It is even more popular option than ZX-100(S), which was created mainly for ZX-C+. The main difference between VX11 and ZX-100 is the GPU: Chrome 640/645 [1106:3a01] vs Chrome 320 [1d17:3a02] (or rather C-320, since they can't use the Chrome brand, because VIA sold S3G to HTC). Both are based on the Excalibur microarchitecture.
- Chrome 640 (C-640) iGPU & Chrome 645 (C-645) iGPU from the VX11/VX11H/VX11PH chipset - Vendor Id: 1106 ("VIA Technologies, Inc."), Device Id: 3a01 ("VX11 Graphics [Chrome 645/640]")
- Chrome 320 (C-320) iGPU from the ZX-100S chipset - Vendor Id: 1d17 ("Zhaoxin"), Device Id: 3a02 ("ZX-100 C-320 GPU")
- Chrome 860 (C-860) iGPU from the ZX-D/KX-5000 (codename "WuDaoKou") SoC - Vendor Id: 1d17 ("Zhaoxin"), Device Id: 3a03 ("ZX-D C-860 GPU")
- Chrome 960 (C-960) iGPU from the ZX-E/KX-6000 (codename "LuJiaZui") SoC - Vendor Id: 1d17 ("Zhaoxin"), Device Id: 3a04 ("ZX-E C-960 GPU")

I have confirmation that these motherboards has the VIA VX11 chipset and the ZX-C+ CPU:
http://archive.is/tiaPR/c17dc7beb984a3028586971f40600a40be1dfc24.jpg
http://archive.is/tiaPR/97120ab0f645a8fcceeb55d3345dc2b21d43629c.jpg
http://archive.is/tiaPR/1fe2d100322d48e1042687a71c086a2ce25755dc.jpg
http://archive.is/tiaPR/b14dc77bc8ac974b3dd9dd1876fc76ac150089b3.jpg
https://i.ibb.co/6BgMwFn/C4710-photo1.png
https://i.ibb.co/xmJC2FG/C4710-photo2.png
https://i.ibb.co/m0NvRkn/C4710-photo3.png
https://i.ibb.co/3fH4Ns8/C4710-photo4.png
As you can see, there is visible model string: 90EWZ1G4CN

See also:
https://www.forum-3dcenter.org/vbulletin/showthread.php?p=12203741#post12203741


The cores support instruction set extensions such as SSE 4.2 and AVX


Die Kerne unterstützen Befehlssatzerweiterungen wie SSE 4.2 und AVX

As far as I know, even CNR has support for AVX2, at least when it comes to these instructions:
https://pastebin.com/raw/4TRjNZkT

However, it should be noted that CNR doesn't support FMA3.
https://www.forum-3dcenter.org/vbulletin/showthread.php?p=11018020#post11018020

I've seen that one about the VIA C4650. It can't run the AVX2 binary because it needs more than that. You need AVX, AVX2, FMA3, BMI2, and ABM. Haswell and Excavator have all of them. I believe the VIA is missing the last 3.

So the naming may be a bit deceptive. Generally speaking these instructions sets are linear in that a new one implies all the old ones and the ones that come out simultaneously on the same generation. AVX2, and BMI2 all came out at the same time on Haswell. FMA3 came slightly earlier with Piledriver. And all processors with AVX2 have FMA3. So without knowledge of the VIA processor, everyone who targets Haswell will use all of them.

The good thing is that y-cruncher actually checks for everything before using the binary. And it correctly detected the lack of FMA3 and BMI2 and decided not to use the AVX2 binary. Otherwise, it would just crash.


See also:
https://raw.githubusercontent.com/InstLatx64/InstLatx64/e0512f539864ab66f7bc311ecfbb4ae3b0ebee44/CentaurHauls/CentaurHauls00006FE_CNR_Isaiah_CPUID.png
https://raw.githubusercontent.com/InstLatx64/InstLatx64/e0512f539864ab66f7bc311ecfbb4ae3b0ebee44/CentaurHauls/CentaurHauls00006FE_CNR_Isaiah_CPUID3.txt
https://raw.githubusercontent.com/InstLatx64/InstLatx64/e0512f539864ab66f7bc311ecfbb4ae3b0ebee44/CentaurHauls/CentaurHauls00006FE_CNR_Isaiah_InstLatX64.txt


Bei der iGPU handelt es sich um eine C-960, die zumindest laut Zhaoxin auf einer eigenen Lösung statt auf einer Chrome von HTCs S3 Graphics basiert.


The iGPU is a C-960, which, at least according to Zhaoxin, is based on its own solution instead of a Chrome from HTC's S3 Graphics.

It's not true. It still based on S3G's IP. What's more, it uses the same microarchitecture as Chrome 640/645, C-320 or C-860: Excalibur.

Here you can find my documentation of C-860 AKA CHX001:
https://www.forum-3dcenter.org/vbulletin/showthread.php?p=12205314#post12205314

C-960 is only a slightly improvement over C-860.


Sie unterstützt OpenCL 1.1, OpenGL 3.2 und Direct3D 11.


It supports OpenCL 1.1, OpenGL 3.2 and Direct3D 11.

It has support for OpenGL ES 2.0 as well. According to Zhaoxin, it also supports OpenVG 1.1 (please note that back in time VIA/S3G claimed to provide an OpenVG 1.1 driver for Destination GPUs, including Chrome 5400E and eH1). What's more, it should be possible to use 3D acceleration via Direct3D 12 API (feature level: Direct3D 11 - "DX11 with DX12 API support").

See also:
https://www.forum-3dcenter.org/vbulletin/attachment.php?attachmentid=67577&d=1566792950


Hier zeigt sich jedoch ein weiteres Problem mit dem Chip, denn für die integrierte GPU auf Basis einer S3-Grafikeinheit gibt es keinen Treiber.


However, this shows another problem with the chip, because there is no driver for the integrated GPU based on an S3 graphics unit.

Of course Linux drivers exist. I have access to the following drivers:
- zhaoxin-chx001-linux-graphics-driver-19.00.33_x86_64.deb (SHA2: 53d591eb907791e8ac035aea8c7a6982a72c17b37e763c8200517112394b49f5) - DRIVER_DATE: 2017-10-30, upload date: 2018-01-10
- zhaoxin-chx001-linux-graphics-driver-19.00.34_x86_64.deb (SHA2: 763c045ee8ab68e95d545278cb3aca75308553cc444afe7d600f839cee2d2604) - DRIVER_DATE: 2017-11-07, upload date: 2018-01-10
- CHX001 中科方德3.0x64_19.00.38.zip (SHA2: 6dac1e3127f0dfc16683ca767322f55dd1c1d07242a5373e2cdcadb18f0c2f34) - DRIVER_DATE: 2017-12-28, upload date: 2018-01-10
- zhaoxin-graphics-19.00.38-i386.tar.bz2 (SHA2: 6d7edd8acec49d1ef22a8bc64250274f08e46b2388d21d8848cc3ed55ecab16d) - DRIVER_DATE: 2017-12-28, upload date: 2018-01-10
- zhaoxin-chx001-linux-graphics-driver-19.00.38_i386.deb (SHA2: 9d939ef84c4e59c18e9de70e91d31bccbf487a837073477be1a55d7577334c2d) - DRIVER_DATE: 2017-12-28, upload date: 2018-01-10
- zhaoxin-graphics-19.00.38-x86_64.tar.bz2 (SHA2: ab83a6eeed977e08eedda403d808ed7cf4d2aa32e5546114f9cdfcd8d7ddba4e) - DRIVER_DATE: 2017-12-28, upload date: 2018-01-10
- zhaoxin-chx001-linux-graphics-driver-19.00.38_x86_64.deb (SHA2: 34458e6b71df181a9006aa871806475fafa3eb2f8ddc9973207aa1105771cd38) - DRIVER_DATE: 2017-12-28, upload date: 2018-01-10
- CHX001 中科方德3.0x64-GFX-19.00.40.zip (SHA2: db1d61228814c9922e1865ef1c02c79b71976d399723be6db3d94b86efae6155) - DRIVER_DATE: 2018-01-26, upload date: 2018-03-05
- zhaoxin-graphics-19.00.40-i386.tar.bz2 (SHA2: 32b1f50e9713c684be8947097733ace0395e2f587b98a52916e9cb9d0d0a1b03) - DRIVER_DATE: 2018-01-26, upload date: 2018-03-05
- zhaoxin-chx001-linux-graphics-driver-19.00.40_i386.deb (SHA2: 7a887d7cfe0c07dd48243a5ad08705d1bbb93e41be0fa329d7c0556ce50163f7) - DRIVER_DATE: 2018-01-26, upload date: 2018-03-05
- zhaoxin-graphics-19.00.40-x86_64.tar.bz2 (SHA2: d3eee7cdbe81c39269819a3450c5f668f4b710581c8741ee5f1ff7fa5ae268c8) - DRIVER_DATE: 2018-01-26, upload date: 2018-03-05
- zhaoxin-chx001-linux-graphics-driver-19.00.40_x86_64.deb (SHA2: 909d70c6314a34cb4fda3c830e607721b25afff715236806955ec0916fe39830) - DRIVER_DATE: 2018-01-26, upload date: 2018-03-05
- ZX-D 中科方德3.0_x64_GFX_19.00.45.zip (SHA2: b6df0329c3f451f8d9bcf41dcf85effc477682c3e7d136355211dd22cfd4040b) - DRIVER_DATE: 2018-05-02, upload date: 2018-05-11
- zhaoxin-ZX-D-graphics-19.00.45-i386.tar.bz2 (SHA2: 71e73d9ecae21cc1fb45d702884e1cc59f89b4b7cdb1ffa5d0c31d0f9cce9f67) - DRIVER_DATE: 2018-05-02, upload date: 2018-05-11
- zhaoxin-ZX-D-linux-graphics-driver-19.00.45_i386.deb (SHA2: 5d464cfdaa5fde0151954cdbbed12b2fe2194e98ce5813ededd1e52111781fd4) - DRIVER_DATE: 2018-05-02, upload date: 2018-05-11
- zhaoxin-ZX-D-graphics-19.00.45-x86_64.tar.bz2 (SHA2: b1651f78ae5eb09b731afa0abf85866a4f941506fd98f31f5cb12f3d85dc1b09) - DRIVER_DATE: 2018-05-02, upload date: 2018-05-11
- zhaoxin-ZX-D-linux-graphics-driver-19.00.45_x86_64.deb (SHA2: b1d9a6d8b622caba0b1aa3cc3e2a82c68491659fa96a47e0066e6046f1e4dba0) - DRIVER_DATE: 2018-05-02, upload date: 2018-05-11
- zhaoxin-chx001-linux-graphics-driver_19.00.45_amd64.deb (SHA2: 63291f86cfd58e40048d92480e89619673882da2222939cb163f57730fb5e3f2) - DRIVER_DATE: 2018-05-02, upload date: 2018-05-11
- zhaoxin-chx001-linux-graphics-driver_19.00.45+1nfs1_amd64.deb (SHA2: 982284fdebe368906223ed282cf41377a917e030ae3eb163024df6c9a3c2e5a2) - DRIVER_DATE: 2018-05-02, upload date: 2018-05-30
- ZX-D+中科方德3.0_x64_19.00.46.zip (SHA2: a070819ecd5cb58d14f4a5b2fbf6d39293efd454ceefabe459caa9e5ad5b8ec8) - DRIVER_DATE: 2018-05-16, upload date: 2018-06-22
- zhaoxin-ZX-D-graphics-19.00.46-i386.tar.bz2 (SHA2: c57ec559919836319131daab01ca87771dcecab5bed76b16aeb10662d59fae9d) - DRIVER_DATE: 2018-05-16, upload date: 2018-06-22
- zhaoxin-ZX-D-linux-graphics-driver-19.00.46_i386.deb (SHA2: 54c6f681397ac3286baa9ea62a06357d98848b8b15b1f842c1e54cf3d4dc22a6) - DRIVER_DATE: 2018-05-16, upload date: 2018-06-22
- zhaoxin-ZX-D-graphics-19.00.46-x86_64.tar.bz2 (SHA2: 6676d7fc3a879c1201762829afb8a088ec671987fe3a854b6adfe8e923a548d4) - DRIVER_DATE: 2018-05-16, upload date: 2018-06-22
- zhaoxin-ZX-D-linux-graphics-driver-19.00.46_x86_64.deb (SHA2: 4be505a20a3fb7e230ddab6c6c60780574e8fb317bc11361c8057a90f7954bbd) - DRIVER_DATE: 2018-05-16, upload date: 2018-06-22
- 19.00.47b.zip (SHA2: 045fa9c8ec4cd3279eb35737b4dc1881da1f784214c0e6cc8c110688ac626d01) - DRIVER_DATE: 2018-07-31, upload date: 2018-08-10
- zhaoxin-ZX-D-linux-graphics-driver-19.00.47b_x86_64.deb (SHA2: f4ada6ca5679419456a3fe1c7e4b82af0414d3058e06651be2258256a52d1435) - DRIVER_DATE: 2018-07-31, upload date: 2018-08-10
- zhaoxin-chx001-linux-graphics-driver_19.00.47b_amd64.deb (SHA2: 2d960d349ca5ee57e62fb8ea303064b2df49f6f3e58b235812dee1c98ac2b7c7) - DRIVER_DATE: 2018-07-31, upload date: 2018-08-10
- zhaoxin-ZX-D-linux-graphics-driver-19.00.47c_x86_64.deb (SHA2: b4f63131053106cc5ad0df082da4f34e0acaefffe200b3eb7e1ac1660bc3681d) - DRIVER_DATE: 2018-10-10, upload date: 2018-10-12
- zhaoxin-chx001-linux-graphics-driver_19.00.47c_amd64.deb (SHA2: 172495e7e953d1a3f2367967959873daea87e0564a8a5b9e1e00dd749ca8cbd9) - DRIVER_DATE: 2018-10-10, upload date: 2018-10-12
- zhaoxin-linux-graphics-driver-dri_21.00.13_amd64.deb (SHA2: dc339abac231df1362c2765f9f97f330311c4232c1d0173ac2dc32d9ed031096) - DRIVER_DATE: 2019-12-12, upload date: 2019-12-14
- zhaoxin-linux-graphics-driver-dri_21.00.16_amd64.deb (SHA2: e2cbd6262af696aac657c18bb810034de22d2dd6cc5c518bd4260744240ecff0) - DRIVER_DATE: 2020-02-25, upload date: 2020-03-21

The latest 21.00.16 driver ("zhaoxin-linux-graphics-driver-dri_21.00.16_amd64.deb" from 2020-03-21 according to upload date, or 2020-02-25 according to DRIVER_DATE) provides support for:
- Chrome 860 (C-860) iGPU from the ZX-D/KX-5000 (codename "WuDaoKou") SoC - Vendor Id: 1d17 ("Zhaoxin"), Device Id: 3a03 ("ZX-D C-860 GPU")
- Chrome 960 (C-960) iGPU from the ZX-E/KX-6000 (codename "LuJiaZui") SoC - Vendor Id: 1d17 ("Zhaoxin"), Device Id: 3a04 ("ZX-E C-960 GPU")


static struct pci_device_id pciidlist[] =
{
{0x1d17, 0x3A03, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0}, //chx001
{0x1d17, 0x3A04, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0}, //chx002 qt
{0, 0, 0}
};


The old 19.00.47c driver ("zhaoxin-ZX-D-linux-graphics-driver-19.00.47c_x86_64.deb"/"zhaoxin-chx001-linux-graphics-driver_19.00.47c_amd64.deb" from 2018-10-12 according to upload date, or 2018-10-10 according to DRIVER_DATE) provided support for:
- Chrome 640 (C-640) iGPU & Chrome 645 (C-645) iGPU from the VX11/VX11H/VX11PH chipset - Vendor Id: 1106 ("VIA Technologies, Inc."), Device Id: 3a01 ("VX11 Graphics [Chrome 645/640]")
- Chrome 860 (C-860) iGPU from the ZX-D/KX-5000 (codename "WuDaoKou") SoC - Vendor Id: 1d17 ("Zhaoxin"), Device Id: 3a03 ("ZX-D C-860 GPU")


static struct pci_device_id pciidlist[] =
{
{0x1106, 0x3A01, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0},
{0x1106, 0x3A03, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0}, //chx001 qt
{0x1d17, 0x3A03, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0}, //chx001
{0, 0, 0}
};


We can assume that pre-19 series driver had support for older Zhaoxin's GPUs:
- Chrome 320 (C-320) iGPU from the ZX-100S chipset - Vendor Id: 1d17 ("Zhaoxin"), Device Id: 3a02 ("ZX-100 C-320 GPU")

Please keep in mind that the standard driver package doesn't include stuff related to OpenCL. It requires an additional driver.

Major issues with the current drivers:
- Although these drivers are included in Chinese distributions (e.g. UOS or NFS China), there are no repos, nor packages for "Western" Linux distros, such as Fedora or openSUSE.
- Support for hybrid GPU configurations requires a patched version of Mesa.
- Support for older GPUs is relatively quickly dropped: 19-series driver doesn't support Chrome 320 from the ZX-100(S) chipset, 21-series driver doesn't support Chrome 640/645 from the VX11(PH) chipset - in other words, the latest driver supports only ZX-D (2017) and ZX-E (2018) hardware.
- OpenGL support is outdated: no support for OGL 4.x at all.
- OpenGL ES support is outdated: no support for OGL ES 3.x at all.
- OpenCL support is outdated: no support for OCL 2.x at all.
- No support for Vulkan at all.
- Although VDPAU and VA-API are supported, it still requires a modified version of MPlayer/SMPlayer to make use of video decoding acceleration.
- Video encoding for H.264/H.265 is limited to 1920x1080p50.
- 3D performance is comparable to the Chrome 640/645 GPU from 2012.

See also:
https://www.forum-3dcenter.org/vbulletin/showthread.php?p=12205314#post12205314


Zwar gibt es und gab es in der Vergangenheit immer wieder Versuche der Linux-Community dahingehend, aber wirklich weit fortgeschritten sind diese nicht.


There have been and have always been attempts by the Linux community to do this, but they are not really very advanced.

It's not true. There are open source drivers for:
- Unichrome series [2002] (from CLE266's iGPU to CX700M(2)'s iGPU), which are based on updated ProSavage - hybrid Savage 4/Savage 2000 [2000]: Savage 4 3D core [1998] and Savage 2000 2D functionality [1999]) → pre-DirectX 9 era (DirectX 7 without T&L)
- Chrome9 series [2005] (from K8M890's iGPU to VX900's iGPU), which derived from DeltaChrome - Columbia microarchitecture [2004] → DirectX 9 era
However, literally no one has worked on open source drivers for:
- Chrome 400 series [2008] (from Chrome 430 GT to Chrome 4300E) and 500 series [2009] (from Chrome 540 GTX to VN1000's iGPU) - Destination microarchitecture [2008] → DirectX 10 era
- Chrome 600 series [2012] and derivatives [2016] - Excalibur microarchitecture [2012] → DirectX 11 era
Not a single line of code has been created for this!

In the past, VIA provided proprietary driver for the Chrome9 series ("via_chrome9_dri.so"/"via_drv.so"), and S3G provided proprietary driver for the Chrome 400/500 series ("s3g_dri.so"/"s3g_drv.so").
Currently, Zhaoxin supports the Chrome 600 series and its derivatives ("zx_dri.so"/"zx_drv.so"). This includes Chrome 640/645 from the VX11(PH) chipset, as well as C-860 from the ZX-D/KX-5000 SoC and C-960 from the ZX-E/KX-6000 SoC. Anyway, it is a proprietary driver, similar to the NVIDIA driver (open source "glue" over a binary blob).

See also:
https://www.phoronix.com/forums/forum/linux-graphics-x-org-drivers/x-org-drm/1026981-it-s-still-going-to-be-tough-getting-the-openchrome-via-kms-driver-in-the-linux-kernel?p=1027179#post1027179
https://www.freedesktop.org/wiki/Openchrome/SupportedHardware/


Zhaoxin arbeitet längst an den KX-7000 alias ZX-F für 2021. Die nutzen weiterhin acht Kerne, sollen aber Big- statt Little-Cores sein. Die Chinesen wollen die Leistung pro Takt um mehrere Faktoren steigern und auch einen DDR5-Speichercontroller integrieren, was durch TSMCs 7-nm-Verfahren unterstützt wird.


Zhaoxin has long been working on the KX-7000 alias ZX-F for 2021. They continue to use eight cores, but are said to be big cores instead of little cores. The Chinese want to increase the performance per clock by several factors and also integrate a DDR5 memory controller, which is supported by TSMC's 7 nm process.

Such were their plans, but they have changed them some time ago. What's more, we already have an early samples of ZX-F, produced in 2019:
https://browser.geekbench.com/v5/cpu/526995

Anyway, ZX-F should be produced in the TSMC's 16 nm process and supports DDR4.
ZX-G will continue to support DDR4, but it should be produced in the TSMC's 7 nm process. Moreover, it may provide even 16 cores.
The first generation with support for DDR5 will be probably ZX-H.
Of course, these are just plans and they may change in the future. Zhaoxin confirmed that they will try to use modern technologies available on the market. However, they can't use something that is not there yet, and the release date of DDR5 has been delayed. According to Gamers Nexus, a leaked internal AMD roadmap is reported to show DDR5 support for 2022 Zen 4 CPUs and Zen 3+ APU. As you can see, it not really the near future.

See also:
https://www.forum-3dcenter.org/vbulletin/showthread.php?p=12205567#post12205567
https://www.forum-3dcenter.org/vbulletin/showthread.php?p=12231741#post12231741
https://www.forum-3dcenter.org/vbulletin/attachment.php?attachmentid=69318
http://www.zhaoxin.com/InCenterContent.aspx?id=282
http://www.gdinsa.org/uploadfile/2020/0326/20200326055237805.pdf

[Tralalak]

to scx: I didn't believe I would see it.
Finally someone self-sufficient.
:up::umassa:

[y33H@]

Great work! Unfortunately, I found some inaccuracies.I have to thank you, even with checking some of your former posts multiple times, there were unfortunately (;() still inaccuracies [typos like KX-5000 for ZX-C+ (the table was already correct) and not saying "up to" 8C] - I apologize, and I've corrected them linking your post. In case of one quadcore vs two dualcores on a package, it seems to be lost in translation - I meant, there was a CNQ 2C+2C and later a native Via CNR 4C. For integrated graphics I willfully wrote "according to Zhaoxing", but obviously all hints go for S3 Chrome. Regarding Linux I'll talk to my Linux guy to update the review.

I love this thread :biggrin:

[Tralalak]

Centaur Technology CHA001 MB Default string
CentaurHauls CPU @2000 MHz (8C 2GHz, 8x 256kB L2, 16MB L3)
source: https://ranker.sisoftware.co.uk/show_device.php?q=c9a598dbbed0a4c5b0c28aeb9ef281a193a393a385e2dff2c3e597aa9bbdd4 e9d9ff97aa9abcc4f9c9ef8aefd2e2c4b78ab2&l=en

[scx]

Zhaoxin's drivers in UOS 20


/data/var/lib/dkms/zx/21.00.13/4.19.0-6-amd64/x86_64/modules/zx.ko
/etc/modprobe.d/zx_gfx.conf
/etc/X11/xorg.conf.d/10-zxgpu.conf
/usr/lib/modules/4.19.0-6-amd64/updates/zx.ko
/usr/lib/x86_64-linux-gnu/dri/ZXEApp.cfg
/usr/lib/x86_64-linux-gnu/dri/zx_dri.so
/usr/lib/x86_64-linux-gnu/dri/zx_drv_video.so
/usr/lib/x86_64-linux-gnu/libkeinterface_zx.so.0.0.0
/usr/lib/x86_64-linux-gnu/libvdpau_zx.so
/usr/lib/xorg/modules/drivers/zx_drv.so
/usr/local/bin/zx_print_driver_map_list
/usr/local/bin/zx_print_driver_map_list_x86
/usr/share/X11/xorg.conf.d/10-zxgpu.conf
/usr/src/zx-21.00.13/built-in_x86_64.o_shipped
/usr/src/zx-21.00.13/cbios/Callback/CBiosCallbacks.c
/usr/src/zx-21.00.13/cbios/Callback/CBiosCallbacks.h
/usr/src/zx-21.00.13/cbios/Callback/MAKEFILE
/usr/src/zx-21.00.13/cbios/Callback/sources
/usr/src/zx-21.00.13/cbios/CBios.h
/usr/src/zx-21.00.13/cbios/Device/CBiosChipShare.h
/usr/src/zx-21.00.13/cbios/Device/CBiosCompile.h
/usr/src/zx-21.00.13/cbios/Device/CBiosDevice.c
/usr/src/zx-21.00.13/cbios/Device/CBiosDevice.h
/usr/src/zx-21.00.13/cbios/Device/CBiosDeviceShare.c
/usr/src/zx-21.00.13/cbios/Device/CBiosDeviceShare.h
/usr/src/zx-21.00.13/cbios/Device/CBiosReg.h
/usr/src/zx-21.00.13/cbios/Device/CBiosShare.c
/usr/src/zx-21.00.13/cbios/Device/CBiosShare.h
/usr/src/zx-21.00.13/cbios/Device/CBiosTypes.h
/usr/src/zx-21.00.13/cbios/Device/CBIOSVER.H
/usr/src/zx-21.00.13/cbios/Device/dirs
/usr/src/zx-21.00.13/cbios/Device/gcc_stdarg.h
/usr/src/zx-21.00.13/cbios/Device/Monitor/CBiosCRTMonitor.c
/usr/src/zx-21.00.13/cbios/Device/Monitor/CBiosCRTMonitor.h
/usr/src/zx-21.00.13/cbios/Device/Monitor/CBiosDPMonitor.c
/usr/src/zx-21.00.13/cbios/Device/Monitor/CBiosDPMonitor.h
/usr/src/zx-21.00.13/cbios/Device/Monitor/CBiosEDPPanel.c
/usr/src/zx-21.00.13/cbios/Device/Monitor/CBiosEDPPanel.h
/usr/src/zx-21.00.13/cbios/Device/Monitor/CBiosHDMIMonitor.c
/usr/src/zx-21.00.13/cbios/Device/Monitor/CBiosHDMIMonitor.h
/usr/src/zx-21.00.13/cbios/Device/Monitor/dirs
/usr/src/zx-21.00.13/cbios/Device/Monitor/DSIPanel/CBiosDSIPanel.c
/usr/src/zx-21.00.13/cbios/Device/Monitor/DSIPanel/CBiosDSIPanel.h
/usr/src/zx-21.00.13/cbios/Device/Monitor/DSIPanel/CBiosHX8392A.c
/usr/src/zx-21.00.13/cbios/Device/Monitor/DSIPanel/CBiosNT35595.c
/usr/src/zx-21.00.13/cbios/Device/Monitor/DSIPanel/CBiosR63319.c
/usr/src/zx-21.00.13/cbios/Device/Monitor/DSIPanel/CBiosR63417.c
/usr/src/zx-21.00.13/cbios/Device/Monitor/DSIPanel/MAKEFILE
/usr/src/zx-21.00.13/cbios/Device/Monitor/DSIPanel/sources
/usr/src/zx-21.00.13/cbios/Device/Monitor/EDPPanel/CBiosITN156.c
/usr/src/zx-21.00.13/cbios/Device/Monitor/EDPPanel/MAKEFILE
/usr/src/zx-21.00.13/cbios/Device/Monitor/EDPPanel/sources
/usr/src/zx-21.00.13/cbios/Device/Port/CBiosCRT.c
/usr/src/zx-21.00.13/cbios/Device/Port/CBiosCRT.h
/usr/src/zx-21.00.13/cbios/Device/Port/CBiosDP.c
/usr/src/zx-21.00.13/cbios/Device/Port/CBiosDP.h
/usr/src/zx-21.00.13/cbios/Device/Port/CBiosDSI.c
/usr/src/zx-21.00.13/cbios/Device/Port/CBiosDSI.h
/usr/src/zx-21.00.13/cbios/Device/Port/CBiosDVO.c
/usr/src/zx-21.00.13/cbios/Device/Port/CBiosDVO.h
/usr/src/zx-21.00.13/cbios/Device/Port/MAKEFILE
/usr/src/zx-21.00.13/cbios/Device/Port/sources
/usr/src/zx-21.00.13/cbios/Display/CBiosDisplayManager.c
/usr/src/zx-21.00.13/cbios/Display/CBiosDisplayManager.h
/usr/src/zx-21.00.13/cbios/Display/CBiosMode.c
/usr/src/zx-21.00.13/cbios/Display/CBiosMode.h
/usr/src/zx-21.00.13/cbios/Display/CBiosPathManager.c
/usr/src/zx-21.00.13/cbios/Display/CBiosPathManager.h
/usr/src/zx-21.00.13/cbios/Display/MAKEFILE
/usr/src/zx-21.00.13/cbios/Display/sources
/usr/src/zx-21.00.13/cbios/Hw/CBiosChipFunc.c
/usr/src/zx-21.00.13/cbios/Hw/CBiosChipFunc.h
/usr/src/zx-21.00.13/cbios/Hw/CBiosHwShare.h
/usr/src/zx-21.00.13/cbios/Hw/CBiosPWM.c
/usr/src/zx-21.00.13/cbios/Hw/CBiosPWM.h
/usr/src/zx-21.00.13/cbios/Hw/Chx001/CBiosVCP_chx.c
/usr/src/zx-21.00.13/cbios/Hw/Chx001/CBiosVCP_chx.h
/usr/src/zx-21.00.13/cbios/Hw/Chx001/CBios_chx.c
/usr/src/zx-21.00.13/cbios/Hw/Chx001/CBios_chx.h
/usr/src/zx-21.00.13/cbios/Hw/Chx001/MAKEFILE
/usr/src/zx-21.00.13/cbios/Hw/Chx001/sources
/usr/src/zx-21.00.13/cbios/Hw/dirs
/usr/src/zx-21.00.13/cbios/Hw/HwBlock/CBiosDIU_CRT.c
/usr/src/zx-21.00.13/cbios/Hw/HwBlock/CBiosDIU_CRT.h
/usr/src/zx-21.00.13/cbios/Hw/HwBlock/CBiosDIU_CSC.c
/usr/src/zx-21.00.13/cbios/Hw/HwBlock/CBiosDIU_CSC.h
/usr/src/zx-21.00.13/cbios/Hw/HwBlock/CBiosDIU_DP.c
/usr/src/zx-21.00.13/cbios/Hw/HwBlock/CBiosDIU_DP.h
/usr/src/zx-21.00.13/cbios/Hw/HwBlock/CBiosDIU_DVO.c
/usr/src/zx-21.00.13/cbios/Hw/HwBlock/CBiosDIU_DVO.h
/usr/src/zx-21.00.13/cbios/Hw/HwBlock/CBiosDIU_HDAC.c
/usr/src/zx-21.00.13/cbios/Hw/HwBlock/CBiosDIU_HDAC.h
/usr/src/zx-21.00.13/cbios/Hw/HwBlock/CBiosDIU_HDCP.c
/usr/src/zx-21.00.13/cbios/Hw/HwBlock/CBiosDIU_HDCP.h
/usr/src/zx-21.00.13/cbios/Hw/HwBlock/CBiosDIU_HDMI.c
/usr/src/zx-21.00.13/cbios/Hw/HwBlock/CBiosDIU_HDMI.h
/usr/src/zx-21.00.13/cbios/Hw/HwBlock/CBiosDIU_HDTV.c
/usr/src/zx-21.00.13/cbios/Hw/HwBlock/CBiosDIU_HDTV.h
/usr/src/zx-21.00.13/cbios/Hw/HwBlock/CBiosIGA_Timing.c
/usr/src/zx-21.00.13/cbios/Hw/HwBlock/CBiosIGA_Timing.h
/usr/src/zx-21.00.13/cbios/Hw/HwBlock/CBiosPHY_DP.c
/usr/src/zx-21.00.13/cbios/Hw/HwBlock/CBiosPHY_DP.h
/usr/src/zx-21.00.13/cbios/Hw/HwBlock/CBiosScaler.c
/usr/src/zx-21.00.13/cbios/Hw/HwBlock/CBiosScaler.h
/usr/src/zx-21.00.13/cbios/Hw/HwBlock/MAKEFILE
/usr/src/zx-21.00.13/cbios/Hw/HwBlock/sources
/usr/src/zx-21.00.13/cbios/Hw/HwCallback/CBiosCallbacksHw.c
/usr/src/zx-21.00.13/cbios/Hw/HwCallback/CBiosCallbacksHw.h
/usr/src/zx-21.00.13/cbios/Hw/HwCallback/MAKEFILE
/usr/src/zx-21.00.13/cbios/Hw/HwCallback/sources
/usr/src/zx-21.00.13/cbios/Hw/HwInit/CBiosInitHw.c
/usr/src/zx-21.00.13/cbios/Hw/HwInit/CBiosInitHw.h
/usr/src/zx-21.00.13/cbios/Hw/HwInit/MAKEFILE
/usr/src/zx-21.00.13/cbios/Hw/HwInit/sources
/usr/src/zx-21.00.13/cbios/Hw/HwInterface/CBiosHwInterface.c
/usr/src/zx-21.00.13/cbios/Hw/HwInterface/CBiosHwInterface.h
/usr/src/zx-21.00.13/cbios/Hw/HwInterface/MAKEFILE
/usr/src/zx-21.00.13/cbios/Hw/HwInterface/sources
/usr/src/zx-21.00.13/cbios/Hw/HwUtil/CBiosI2C.c
/usr/src/zx-21.00.13/cbios/Hw/HwUtil/CBiosI2C.h
/usr/src/zx-21.00.13/cbios/Hw/HwUtil/CBiosUtilHw.c
/usr/src/zx-21.00.13/cbios/Hw/HwUtil/CBiosUtilHw.h
/usr/src/zx-21.00.13/cbios/Hw/HwUtil/MAKEFILE
/usr/src/zx-21.00.13/cbios/Hw/HwUtil/sources
/usr/src/zx-21.00.13/cbios/Hw/Interrupt/CBiosISR.c
/usr/src/zx-21.00.13/cbios/Hw/Interrupt/MAKEFILE
/usr/src/zx-21.00.13/cbios/Hw/Interrupt/sources
/usr/src/zx-21.00.13/cbios/Hw/Register/BIU_CR_C_BUS_registers.h
/usr/src/zx-21.00.13/cbios/Hw/Register/BIU_HDA_registers.h
/usr/src/zx-21.00.13/cbios/Hw/Register/BIU_MM_registers.h
/usr/src/zx-21.00.13/cbios/Hw/Register/BIU_SBI_registers.h
/usr/src/zx-21.00.13/cbios/Hw/Register/BIU_TSR_registers.h
/usr/src/zx-21.00.13/cbios/Hw/Register/BIU_VCP_registers.h
/usr/src/zx-21.00.13/cbios/Hw/Register/DIU_CR_registers.h
/usr/src/zx-21.00.13/cbios/Hw/Register/DIU_MM_registers.h
/usr/src/zx-21.00.13/cbios/Hw/Register/DIU_SR_registers.h
/usr/src/zx-21.00.13/cbios/Hw/Register/DIU_vga_registers.h
/usr/src/zx-21.00.13/cbios/Hw/Register/Monitor/CBiosDPCDRegister.h
/usr/src/zx-21.00.13/cbios/Hw/Register/pmu_registers.h
/usr/src/zx-21.00.13/cbios/Hw/Register/pmu_registers_e2k.h
/usr/src/zx-21.00.13/cbios/Init/CBiosInit.c
/usr/src/zx-21.00.13/cbios/Init/MAKEFILE
/usr/src/zx-21.00.13/cbios/Init/sources
/usr/src/zx-21.00.13/cbios/Interface/CBios.c
/usr/src/zx-21.00.13/cbios/Util/CBiosEDID.c
/usr/src/zx-21.00.13/cbios/Util/CBiosEDID.h
/usr/src/zx-21.00.13/cbios/Util/CBiosGPIO.h
/usr/src/zx-21.00.13/cbios/Util/CBiosUtil.c
/usr/src/zx-21.00.13/cbios/Util/MAKEFILE
/usr/src/zx-21.00.13/cbios/Util/sources
/usr/src/zx-21.00.13/core_errno.h
/usr/src/zx-21.00.13/core_module.c
/usr/src/zx-21.00.13/ddk_types.h
/usr/src/zx-21.00.13/dkms.conf
/usr/src/zx-21.00.13/kernel_import.h
/usr/src/zx-21.00.13/kernel_interface.h
/usr/src/zx-21.00.13/Makefile
/usr/src/zx-21.00.13/os_interface.c
/usr/src/zx-21.00.13/os_interface.h
/usr/src/zx-21.00.13/os_shared.c
/usr/src/zx-21.00.13/zx.h
/usr/src/zx-21.00.13/zxg.c
/usr/src/zx-21.00.13/zxgfx_trace.h
/usr/src/zx-21.00.13/zxgfx_trace_events.c
/usr/src/zx-21.00.13/zx_atomic.c
/usr/src/zx-21.00.13/zx_atomic.h
/usr/src/zx-21.00.13/zx_backlight.c
/usr/src/zx-21.00.13/zx_backlight.h
/usr/src/zx-21.00.13/zx_cbios.c
/usr/src/zx-21.00.13/zx_cbios.h
/usr/src/zx-21.00.13/zx_cec.c
/usr/src/zx-21.00.13/zx_cec.h
/usr/src/zx-21.00.13/zx_connector.c
/usr/src/zx-21.00.13/zx_crtc.c
/usr/src/zx-21.00.13/zx_crtc.h
/usr/src/zx-21.00.13/zx_debugfs.c
/usr/src/zx-21.00.13/zx_debugfs.h
/usr/src/zx-21.00.13/zx_def.h
/usr/src/zx-21.00.13/zx_device_debug.h
/usr/src/zx-21.00.13/zx_disp.c
/usr/src/zx-21.00.13/zx_disp.h
/usr/src/zx-21.00.13/zx_driver.c
/usr/src/zx-21.00.13/zx_driver.h
/usr/src/zx-21.00.13/zx_drmfb.c
/usr/src/zx-21.00.13/zx_drmfb.h
/usr/src/zx-21.00.13/zx_encoder.c
/usr/src/zx-21.00.13/zx_fbdev.c
/usr/src/zx-21.00.13/zx_fbdev.h
/usr/src/zx-21.00.13/zx_fence.c
/usr/src/zx-21.00.13/zx_fence.h
/usr/src/zx-21.00.13/zx_gem.c
/usr/src/zx-21.00.13/zx_gem.h
/usr/src/zx-21.00.13/zx_gem_debug.h
/usr/src/zx-21.00.13/zx_ioctl.c
/usr/src/zx-21.00.13/zx_ioctl.h
/usr/src/zx-21.00.13/zx_irq.c
/usr/src/zx-21.00.13/zx_irq.h
/usr/src/zx-21.00.13/zx_keymap.c
/usr/src/zx-21.00.13/zx_kms.h
/usr/src/zx-21.00.13/zx_led.c
/usr/src/zx-21.00.13/zx_led.h
/usr/src/zx-21.00.13/zx_pcie.c
/usr/src/zx-21.00.13/zx_plane.c
/usr/src/zx-21.00.13/zx_plane.h
/usr/src/zx-21.00.13/zx_sync.c
/usr/src/zx-21.00.13/zx_sync.h
/usr/src/zx-21.00.13/zx_types.h
/usr/src/zx-21.00.13/zx_version.h
/usr/src/zx-21.00.13/.built-in_x86_64.o.cmd
/var/lib/dkms/zx/21.00.13/4.19.0-6-amd64/x86_64/modules/zx.ko


This list doesn't include additional drivers (OpenCL), nor extra applications (special version of MPlayer or SMPlayer).

Anyway, as you can see, source file have changed slightly since the 19-series driver.
https://www.forum-3dcenter.org/vbulletin/showthread.php?p=12205314#post12205314

Kernel driver sources

S3G's 14-series driver for Destination

Package name: s3g-Chrome4xx_5xx-14.05.02-x86_64-new_kernel.tar.bz2 (SHA2: ecd3eaf720c1dadbd85b9609d96b0b7c1388e227158fe8a200a13cc8c0d5d35d)
Driver date: 2011-06-09
Driver version: 14.05.02
Supported GPUs: up to S3G Chrome 500 series ([5333:9060] S3G Chrome 530 GT, S3G Chrome 540 GTX)


Makefile
os_interface.c
os_interface.h
s3.c
s3.h
s3g_def.h
s3g_driver.c
s3g_driver.h
s3g_fb.c
s3g_fb.h
s3g_ioctl.h
s3g_private.o
s3g_version.h
s3_pcie.c


VIA's 15-series driver for Excalibur

Package name: vx11_ubuntu1204_ubuntu1210_15-00-02b.7z.zip (SHA2: 47927e6df202019585bc257284eccce26ae8ea351a9f7a8a2ffb59880cb5021e), vx11_ubuntu1204_ubuntu1210_15-00-02b.7z (SHA2: 93199826974ada2c180a9ce47f5be1e608ea536de0eb7f50a5d06e7e5bf2f1c6)
Driver date: 2012-12-26
Driver version: 15.00.02
Supported GPUs: up to VIA Chrome 600 series ([1106:3a01] VIA Chrome 640/645)


Makefile
os_interface.c
os_interface.h
s3.c
s3.h
s3g_def.h
s3g_driver.c
s3g_driver.h
s3g_fb.c
s3g_fb.h
s3g_ioctl.h
s3g_private.o
s3g_types.h
s3g_version.h
s3_pcie.c


Zhaoxin's 19-series driver for Excalibur

Package name: zhaoxin-ZX-D-linux-graphics-driver-19.00.47c_x86_64.deb (SHA2: b4f63131053106cc5ad0df082da4f34e0acaefffe200b3eb7e1ac1660bc3681d), zhaoxin-chx001-linux-graphics-driver_19.00.47c_amd64.deb (SHA2: 172495e7e953d1a3f2367967959873daea87e0564a8a5b9e1e00dd749ca8cbd9)
Driver date: 2018-10-10
Driver version: 19.00.47c
Supported GPUs: up to Zhaoxin Chrome 800 series ([1d17:3a03] Zhaoxin Chrome 860)


.built-in_64.o.cmd
built-in_64.o
core_errno.h
core_module.c
kernel_import.h
kernel_interface.h
Makefile
os_interface.c
os_interface.h
os_shared.c
s3.h
s3gfb.h
s3gfx_trace.h
s3g_buffer_object.h
s3g_cec.h
s3g_debugfs.h
s3g_def.h
s3g_driver.h
s3g_ioctl.h
s3g_ioctl_compat.h
s3g_led.h
s3g_screen.h
s3g_sync.h
s3g_types.h
s3g_version.h
zx1.c
zxfb.c
zxfb_ops.c
zx_buffer_object.c
zx_cec.c
zx_debugfs.c
zx_driver.c
zx_ioctl.c
zx_ioctl_compat.c
zx_keymap.c
zx_led.c
zx_pcie.c
zx_screen.c
zx_sync.c


Zhaoxin's 21-series driver for Excalibur

Package name: zhaoxin-linux-graphics-driver-dri_21.00.16_amd64.deb (SHA2: e2cbd6262af696aac657c18bb810034de22d2dd6cc5c518bd4260744240ecff0)
Driver date: 2020-02-25
Driver version: 21.00.16
Supported GPUs: up to Zhaoxin Chrome 900 series ([1d17:3a04] Zhaoxin Chrome 960)


.built-in_x86_64.o.cmd
built-in_x86_64.o_shipped
cbios/Callback/CBiosCallbacks.c
cbios/Callback/CBiosCallbacks.h
cbios/Callback/MAKEFILE
cbios/Callback/sources
cbios/CBios.h
cbios/Device/CBiosChipShare.h
cbios/Device/CBiosCompile.h
cbios/Device/CBiosDevice.c
cbios/Device/CBiosDevice.h
cbios/Device/CBiosDeviceShare.c
cbios/Device/CBiosDeviceShare.h
cbios/Device/CBiosReg.h
cbios/Device/CBiosShare.c
cbios/Device/CBiosShare.h
cbios/Device/CBiosTypes.h
cbios/Device/CBIOSVER.H
cbios/Device/dirs
cbios/Device/gcc_stdarg.h
cbios/Device/Monitor/CBiosCRTMonitor.c
cbios/Device/Monitor/CBiosCRTMonitor.h
cbios/Device/Monitor/CBiosDPMonitor.c
cbios/Device/Monitor/CBiosDPMonitor.h
cbios/Device/Monitor/CBiosEDPPanel.c
cbios/Device/Monitor/CBiosEDPPanel.h
cbios/Device/Monitor/CBiosHDMIMonitor.c
cbios/Device/Monitor/CBiosHDMIMonitor.h
cbios/Device/Monitor/dirs
cbios/Device/Monitor/DSIPanel/CBiosDSIPanel.c
cbios/Device/Monitor/DSIPanel/CBiosDSIPanel.h
cbios/Device/Monitor/DSIPanel/CBiosHX8392A.c
cbios/Device/Monitor/DSIPanel/CBiosNT35595.c
cbios/Device/Monitor/DSIPanel/CBiosR63319.c
cbios/Device/Monitor/DSIPanel/CBiosR63417.c
cbios/Device/Monitor/DSIPanel/MAKEFILE
cbios/Device/Monitor/DSIPanel/sources
cbios/Device/Monitor/EDPPanel/CBiosITN156.c
cbios/Device/Monitor/EDPPanel/MAKEFILE
cbios/Device/Monitor/EDPPanel/sources
cbios/Device/Port/CBiosCRT.c
cbios/Device/Port/CBiosCRT.h
cbios/Device/Port/CBiosDP.c
cbios/Device/Port/CBiosDP.h
cbios/Device/Port/CBiosDSI.c
cbios/Device/Port/CBiosDSI.h
cbios/Device/Port/CBiosDVO.c
cbios/Device/Port/CBiosDVO.h
cbios/Device/Port/MAKEFILE
cbios/Device/Port/sources
cbios/Display/CBiosDisplayManager.c
cbios/Display/CBiosDisplayManager.h
cbios/Display/CBiosMode.c
cbios/Display/CBiosMode.h
cbios/Display/CBiosPathManager.c
cbios/Display/CBiosPathManager.h
cbios/Display/MAKEFILE
cbios/Display/sources
cbios/Hw/CBiosChipFunc.c
cbios/Hw/CBiosChipFunc.h
cbios/Hw/CBiosHwShare.h
cbios/Hw/CBiosPWM.c
cbios/Hw/CBiosPWM.h
cbios/Hw/Chx001/CBiosVCP_chx.c
cbios/Hw/Chx001/CBiosVCP_chx.h
cbios/Hw/Chx001/CBios_chx.c
cbios/Hw/Chx001/CBios_chx.h
cbios/Hw/Chx001/MAKEFILE
cbios/Hw/Chx001/sources
cbios/Hw/dirs
cbios/Hw/HwBlock/CBiosDIU_CRT.c
cbios/Hw/HwBlock/CBiosDIU_CRT.h
cbios/Hw/HwBlock/CBiosDIU_CSC.c
cbios/Hw/HwBlock/CBiosDIU_CSC.h
cbios/Hw/HwBlock/CBiosDIU_DP.c
cbios/Hw/HwBlock/CBiosDIU_DP.h
cbios/Hw/HwBlock/CBiosDIU_DVO.c
cbios/Hw/HwBlock/CBiosDIU_DVO.h
cbios/Hw/HwBlock/CBiosDIU_HDAC.c
cbios/Hw/HwBlock/CBiosDIU_HDAC.h
cbios/Hw/HwBlock/CBiosDIU_HDCP.c
cbios/Hw/HwBlock/CBiosDIU_HDCP.h
cbios/Hw/HwBlock/CBiosDIU_HDMI.c
cbios/Hw/HwBlock/CBiosDIU_HDMI.h
cbios/Hw/HwBlock/CBiosDIU_HDTV.c
cbios/Hw/HwBlock/CBiosDIU_HDTV.h
cbios/Hw/HwBlock/CBiosIGA_Timing.c
cbios/Hw/HwBlock/CBiosIGA_Timing.h
cbios/Hw/HwBlock/CBiosPHY_DP.c
cbios/Hw/HwBlock/CBiosPHY_DP.h
cbios/Hw/HwBlock/CBiosScaler.c
cbios/Hw/HwBlock/CBiosScaler.h
cbios/Hw/HwBlock/MAKEFILE
cbios/Hw/HwBlock/sources
cbios/Hw/HwCallback/CBiosCallbacksHw.c
cbios/Hw/HwCallback/CBiosCallbacksHw.h
cbios/Hw/HwCallback/MAKEFILE
cbios/Hw/HwCallback/sources
cbios/Hw/HwInit/CBiosInitHw.c
cbios/Hw/HwInit/CBiosInitHw.h
cbios/Hw/HwInit/MAKEFILE
cbios/Hw/HwInit/sources
cbios/Hw/HwInterface/CBiosHwInterface.c
cbios/Hw/HwInterface/CBiosHwInterface.h
cbios/Hw/HwInterface/MAKEFILE
cbios/Hw/HwInterface/sources
cbios/Hw/HwUtil/CBiosI2C.c
cbios/Hw/HwUtil/CBiosI2C.h
cbios/Hw/HwUtil/CBiosUtilHw.c
cbios/Hw/HwUtil/CBiosUtilHw.h
cbios/Hw/HwUtil/MAKEFILE
cbios/Hw/HwUtil/sources
cbios/Hw/Interrupt/CBiosISR.c
cbios/Hw/Interrupt/MAKEFILE
cbios/Hw/Interrupt/sources
cbios/Hw/Register/BIU_CR_C_BUS_registers.h
cbios/Hw/Register/BIU_HDA_registers.h
cbios/Hw/Register/BIU_MM_registers.h
cbios/Hw/Register/BIU_SBI_registers.h
cbios/Hw/Register/BIU_TSR_registers.h
cbios/Hw/Register/BIU_VCP_registers.h
cbios/Hw/Register/DIU_CR_registers.h
cbios/Hw/Register/DIU_MM_registers.h
cbios/Hw/Register/DIU_SR_registers.h
cbios/Hw/Register/DIU_vga_registers.h
cbios/Hw/Register/Monitor/CBiosDPCDRegister.h
cbios/Hw/Register/pmu_registers.h
cbios/Hw/Register/pmu_registers_e2k.h
cbios/Init/CBiosInit.c
cbios/Init/MAKEFILE
cbios/Init/sources
cbios/Interface/CBios.c
cbios/Util/CBiosEDID.c
cbios/Util/CBiosEDID.h
cbios/Util/CBiosGPIO.h
cbios/Util/CBiosUtil.c
cbios/Util/MAKEFILE
cbios/Util/sources
core_errno.h
core_module.c
ddk_types.h
dkms.conf
kernel_import.h
kernel_interface.h
Makefile
os_interface.c
os_interface.h
os_shared.c
zx.h
zxg.c
zxgfx_trace.h
zxgfx_trace_events.c
zx_atomic.c
zx_atomic.h
zx_backlight.c
zx_backlight.h
zx_cbios.c
zx_cbios.h
zx_cec.c
zx_cec.h
zx_connector.c
zx_crtc.c
zx_crtc.h
zx_debugfs.c
zx_debugfs.h
zx_def.h
zx_device_debug.h
zx_disp.c
zx_disp.h
zx_driver.c
zx_driver.h
zx_drmfb.c
zx_drmfb.h
zx_encoder.c
zx_fbdev.c
zx_fbdev.h
zx_fence.c
zx_fence.h
zx_gem.c
zx_gem.h
zx_gem_debug.h
zx_ioctl.c
zx_ioctl.h
zx_irq.c
zx_irq.h
zx_keymap.c
zx_kms.h
zx_led.c
zx_led.h
zx_pcie.c
zx_plane.c
zx_plane.h
zx_sync.c
zx_sync.h
zx_types.h
zx_version.h


Changes between 19- and 21-series kernel driver sources

Files added


built-in_x86_64.o_shipped
cbios/Callback/CBiosCallbacks.c
cbios/Callback/CBiosCallbacks.h
cbios/Callback/MAKEFILE
cbios/Callback/sources
cbios/CBios.h
cbios/Device/CBiosChipShare.h
cbios/Device/CBiosCompile.h
cbios/Device/CBiosDevice.c
cbios/Device/CBiosDevice.h
cbios/Device/CBiosDeviceShare.c
cbios/Device/CBiosDeviceShare.h
cbios/Device/CBiosReg.h
cbios/Device/CBiosShare.c
cbios/Device/CBiosShare.h
cbios/Device/CBiosTypes.h
cbios/Device/CBIOSVER.H
cbios/Device/dirs
cbios/Device/gcc_stdarg.h
cbios/Device/Monitor/CBiosCRTMonitor.c
cbios/Device/Monitor/CBiosCRTMonitor.h
cbios/Device/Monitor/CBiosDPMonitor.c
cbios/Device/Monitor/CBiosDPMonitor.h
cbios/Device/Monitor/CBiosEDPPanel.c
cbios/Device/Monitor/CBiosEDPPanel.h
cbios/Device/Monitor/CBiosHDMIMonitor.c
cbios/Device/Monitor/CBiosHDMIMonitor.h
cbios/Device/Monitor/dirs
cbios/Device/Monitor/DSIPanel/CBiosDSIPanel.c
cbios/Device/Monitor/DSIPanel/CBiosDSIPanel.h
cbios/Device/Monitor/DSIPanel/CBiosHX8392A.c
cbios/Device/Monitor/DSIPanel/CBiosNT35595.c
cbios/Device/Monitor/DSIPanel/CBiosR63319.c
cbios/Device/Monitor/DSIPanel/CBiosR63417.c
cbios/Device/Monitor/DSIPanel/MAKEFILE
cbios/Device/Monitor/DSIPanel/sources
cbios/Device/Monitor/EDPPanel/CBiosITN156.c
cbios/Device/Monitor/EDPPanel/MAKEFILE
cbios/Device/Monitor/EDPPanel/sources
cbios/Device/Port/CBiosCRT.c
cbios/Device/Port/CBiosCRT.h
cbios/Device/Port/CBiosDP.c
cbios/Device/Port/CBiosDP.h
cbios/Device/Port/CBiosDSI.c
cbios/Device/Port/CBiosDSI.h
cbios/Device/Port/CBiosDVO.c
cbios/Device/Port/CBiosDVO.h
cbios/Device/Port/MAKEFILE
cbios/Device/Port/sources
cbios/Display/CBiosDisplayManager.c
cbios/Display/CBiosDisplayManager.h
cbios/Display/CBiosMode.c
cbios/Display/CBiosMode.h
cbios/Display/CBiosPathManager.c
cbios/Display/CBiosPathManager.h
cbios/Display/MAKEFILE
cbios/Display/sources
cbios/Hw/CBiosChipFunc.c
cbios/Hw/CBiosChipFunc.h
cbios/Hw/CBiosHwShare.h
cbios/Hw/CBiosPWM.c
cbios/Hw/CBiosPWM.h
cbios/Hw/Chx001/CBiosVCP_chx.c
cbios/Hw/Chx001/CBiosVCP_chx.h
cbios/Hw/Chx001/CBios_chx.c
cbios/Hw/Chx001/CBios_chx.h
cbios/Hw/Chx001/MAKEFILE
cbios/Hw/Chx001/sources
cbios/Hw/dirs
cbios/Hw/HwBlock/CBiosDIU_CRT.c
cbios/Hw/HwBlock/CBiosDIU_CRT.h
cbios/Hw/HwBlock/CBiosDIU_CSC.c
cbios/Hw/HwBlock/CBiosDIU_CSC.h
cbios/Hw/HwBlock/CBiosDIU_DP.c
cbios/Hw/HwBlock/CBiosDIU_DP.h
cbios/Hw/HwBlock/CBiosDIU_DVO.c
cbios/Hw/HwBlock/CBiosDIU_DVO.h
cbios/Hw/HwBlock/CBiosDIU_HDAC.c
cbios/Hw/HwBlock/CBiosDIU_HDAC.h
cbios/Hw/HwBlock/CBiosDIU_HDCP.c
cbios/Hw/HwBlock/CBiosDIU_HDCP.h
cbios/Hw/HwBlock/CBiosDIU_HDMI.c
cbios/Hw/HwBlock/CBiosDIU_HDMI.h
cbios/Hw/HwBlock/CBiosDIU_HDTV.c
cbios/Hw/HwBlock/CBiosDIU_HDTV.h
cbios/Hw/HwBlock/CBiosIGA_Timing.c
cbios/Hw/HwBlock/CBiosIGA_Timing.h
cbios/Hw/HwBlock/CBiosPHY_DP.c
cbios/Hw/HwBlock/CBiosPHY_DP.h
cbios/Hw/HwBlock/CBiosScaler.c
cbios/Hw/HwBlock/CBiosScaler.h
cbios/Hw/HwBlock/MAKEFILE
cbios/Hw/HwBlock/sources
cbios/Hw/HwCallback/CBiosCallbacksHw.c
cbios/Hw/HwCallback/CBiosCallbacksHw.h
cbios/Hw/HwCallback/MAKEFILE
cbios/Hw/HwCallback/sources
cbios/Hw/HwInit/CBiosInitHw.c
cbios/Hw/HwInit/CBiosInitHw.h
cbios/Hw/HwInit/MAKEFILE
cbios/Hw/HwInit/sources
cbios/Hw/HwInterface/CBiosHwInterface.c
cbios/Hw/HwInterface/CBiosHwInterface.h
cbios/Hw/HwInterface/MAKEFILE
cbios/Hw/HwInterface/sources
cbios/Hw/HwUtil/CBiosI2C.c
cbios/Hw/HwUtil/CBiosI2C.h
cbios/Hw/HwUtil/CBiosUtilHw.c
cbios/Hw/HwUtil/CBiosUtilHw.h
cbios/Hw/HwUtil/MAKEFILE
cbios/Hw/HwUtil/sources
cbios/Hw/Interrupt/CBiosISR.c
cbios/Hw/Interrupt/MAKEFILE
cbios/Hw/Interrupt/sources
cbios/Hw/Register/BIU_CR_C_BUS_registers.h
cbios/Hw/Register/BIU_HDA_registers.h
cbios/Hw/Register/BIU_MM_registers.h
cbios/Hw/Register/BIU_SBI_registers.h
cbios/Hw/Register/BIU_TSR_registers.h
cbios/Hw/Register/BIU_VCP_registers.h
cbios/Hw/Register/DIU_CR_registers.h
cbios/Hw/Register/DIU_MM_registers.h
cbios/Hw/Register/DIU_SR_registers.h
cbios/Hw/Register/DIU_vga_registers.h
cbios/Hw/Register/Monitor/CBiosDPCDRegister.h
cbios/Hw/Register/pmu_registers.h
cbios/Hw/Register/pmu_registers_e2k.h
cbios/Init/CBiosInit.c
cbios/Init/MAKEFILE
cbios/Init/sources
cbios/Interface/CBios.c
cbios/Util/CBiosEDID.c
cbios/Util/CBiosEDID.h
cbios/Util/CBiosGPIO.h
cbios/Util/CBiosUtil.c
cbios/Util/MAKEFILE
cbios/Util/sources
ddk_types.h
dkms.conf
zxg.c
zxgfx_trace_events.c
zx_atomic.c
zx_atomic.h
zx_backlight.c
zx_backlight.h
zx_cbios.c
zx_cbios.h
zx_connector.c
zx_crtc.c
zx_crtc.h
zx_def.h
zx_device_debug.h
zx_disp.c
zx_disp.h
zx_drmfb.c
zx_drmfb.h
zx_encoder.c
zx_fbdev.c
zx_fbdev.h
zx_fence.c
zx_fence.h
zx_gem.c
zx_gem.h
zx_gem_debug.h
zx_irq.c
zx_irq.h
zx_kms.h
zx_pcie.c
zx_plane.c
zx_plane.h


Files renamed (and modified)


.built-in_64.o.cmd → .built-in_x86_64.o.cmd
s3.h → zx.h
s3gfx_trace.h → zxgfx_trace.h
s3g_cec.h → zx_cec.h
s3g_debugfs.h → zx_debugfs.h
s3g_def.h → zx_def.h
s3g_driver.h → zx_driver.h
s3g_ioctl.h → zx_ioctl.h
s3g_led.h → zx_led.h
s3g_sync.h → zx_sync.h
s3g_types.h → zx_types.h
s3g_version.h → zx_version.h


Files modified


core_module.c
kernel_import.h
kernel_interface.h
Makefile
os_interface.c
os_interface.h
os_shared.c
zx_cec.c
zx_debugfs.c
zx_driver.c
zx_ioctl.c
zx_keymap.c
zx_led.c
zx_pcie.c
zx_sync.c


Files not-modified

core_errno.h


Files removed


s3gfb.h (see: zx_cbios.h)
s3g_buffer_object.h
s3g_ioctl_compat.h
s3g_screen.h
zx1.c (see: zx_pcie.c, zxg.c)
zxfb.c
zxfb_ops.c (see: zx_fbdev.c)
zx_buffer_object.c
zx_ioctl_compat.c
zx_screen.c

[scx]

Copyright

Copyright of the S3G's 14-series driver for Destination


LICENSE:Copyright (c) 2011 S3 Graphics Co., Ltd. All rights reserved.


Copyright of the VIA's 15-series driver for Excalibur


LICENSE:Copyright (c) 2011 S3 Graphics Co., Ltd. All rights reserved.


Copyright of the Zhaoxin's 19-series driver for Excalibur


$ find * -xtype f | sort -fV | xargs -I{} grep -HE 'Copyright.*(19[0-9]{2}|2[012][0-9]{2})' '{}'
core_errno.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
core_errno.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
core_errno.h:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
core_module.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
core_module.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
core_module.c:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
kernel_import.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
kernel_import.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
kernel_import.h:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
kernel_interface.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
kernel_interface.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
kernel_interface.h:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
os_interface.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
os_interface.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
os_interface.c:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
os_interface.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
os_interface.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
os_interface.h:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
os_shared.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
os_shared.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
os_shared.c:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
s3.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
s3.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
s3.h:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
s3gfb.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
s3gfb.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
s3gfb.h:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
s3gfx_trace.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
s3gfx_trace.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
s3gfx_trace.h:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
s3g_buffer_object.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
s3g_buffer_object.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
s3g_buffer_object.h:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
s3g_cec.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
s3g_cec.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
s3g_cec.h:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
s3g_debugfs.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
s3g_debugfs.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
s3g_debugfs.h:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
s3g_def.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
s3g_def.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
s3g_def.h:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
s3g_driver.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
s3g_driver.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
s3g_driver.h:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
s3g_ioctl.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
s3g_ioctl.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
s3g_ioctl.h:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
s3g_ioctl_compat.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
s3g_ioctl_compat.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
s3g_ioctl_compat.h:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
s3g_led.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
s3g_led.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
s3g_led.h:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
s3g_screen.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
s3g_screen.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
s3g_screen.h:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
s3g_sync.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
s3g_sync.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
s3g_sync.h:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
s3g_types.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
s3g_types.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
s3g_types.h:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
zx1.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
zx1.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
zx1.c:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
zxfb.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
zxfb.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
zxfb.c:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
zxfb_ops.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
zxfb_ops.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
zxfb_ops.c:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
zx_buffer_object.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
zx_buffer_object.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
zx_buffer_object.c:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
zx_cec.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
zx_cec.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
zx_cec.c:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
zx_debugfs.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
zx_debugfs.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
zx_debugfs.c:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
zx_driver.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
zx_driver.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
zx_driver.c:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
zx_ioctl.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
zx_ioctl.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
zx_ioctl.c:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
zx_ioctl_compat.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
zx_ioctl_compat.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
zx_ioctl_compat.c:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
zx_keymap.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
zx_keymap.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
zx_keymap.c:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
zx_led.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
zx_led.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
zx_led.c:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
zx_pcie.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
zx_pcie.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
zx_pcie.c:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
zx_screen.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
zx_screen.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
zx_screen.c:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
zx_sync.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
zx_sync.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
zx_sync.c:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.


Copyright of the Zhaoxin's 21-series driver for Excalibur


$ find * -xtype f | sort -fV | xargs -I{} grep -HE 'Copyright.*(19[0-9]{2}|2[012][0-9]{2})' '{}'
cbios/Callback/CBiosCallbacks.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Callback/CBiosCallbacks.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Callback/CBiosCallbacks.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Callback/CBiosCallbacks.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/CBios.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/CBios.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/CBios.h:* Copyright 2013-2017 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
cbios/Device/CBiosChipShare.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Device/CBiosChipShare.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Device/CBiosCompile.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Device/CBiosCompile.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Device/CBiosDevice.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Device/CBiosDevice.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Device/CBiosDevice.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Device/CBiosDevice.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Device/CBiosDeviceShare.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Device/CBiosDeviceShare.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Device/CBiosDeviceShare.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Device/CBiosDeviceShare.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Device/CBiosReg.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Device/CBiosReg.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Device/CBiosShare.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Device/CBiosShare.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Device/CBiosShare.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Device/CBiosShare.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Device/CBiosShare.h:* Copyright 2013-2017 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
cbios/Device/CBiosTypes.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Device/CBiosTypes.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Device/CBIOSVER.H:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Device/CBIOSVER.H:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Device/CBIOSVER.H:* Copyright 2013-2018 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
cbios/Device/gcc_stdarg.h:/* Copyright (C) 1989, 1997, 1998, 1999, 2000 Free Software Foundation, Inc.
cbios/Device/Monitor/CBiosCRTMonitor.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Device/Monitor/CBiosCRTMonitor.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Device/Monitor/CBiosCRTMonitor.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Device/Monitor/CBiosCRTMonitor.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Device/Monitor/CBiosDPMonitor.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Device/Monitor/CBiosDPMonitor.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Device/Monitor/CBiosDPMonitor.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Device/Monitor/CBiosDPMonitor.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Device/Monitor/CBiosEDPPanel.c:* Copyright 2015 VIA Technologies, Inc. All Rights Reserved.
cbios/Device/Monitor/CBiosEDPPanel.c:* Copyright 2015 S3 Graphics, Inc. All Rights Reserved.
cbios/Device/Monitor/CBiosEDPPanel.c:* Copyright 2015 VIA Alliance(Zhaoxin)Semiconductor Co., Ltd. All Rights Reserved.
cbios/Device/Monitor/CBiosEDPPanel.h:* Copyright 2015 VIA Technologies, Inc. All Rights Reserved.
cbios/Device/Monitor/CBiosEDPPanel.h:* Copyright 2015 S3 Graphics, Inc. All Rights Reserved.
cbios/Device/Monitor/CBiosEDPPanel.h:* Copyright 2015 VIA Alliance(Zhaoxin)Semiconductor Co., Ltd. All Rights Reserved.
cbios/Device/Monitor/CBiosHDMIMonitor.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Device/Monitor/CBiosHDMIMonitor.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Device/Monitor/CBiosHDMIMonitor.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Device/Monitor/CBiosHDMIMonitor.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Device/Monitor/DSIPanel/CBiosDSIPanel.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Device/Monitor/DSIPanel/CBiosDSIPanel.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Device/Monitor/DSIPanel/CBiosDSIPanel.c:* Copyright 2013-2017 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
cbios/Device/Monitor/DSIPanel/CBiosDSIPanel.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Device/Monitor/DSIPanel/CBiosDSIPanel.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Device/Monitor/DSIPanel/CBiosDSIPanel.h:* Copyright 2013-2017 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
cbios/Device/Monitor/DSIPanel/CBiosHX8392A.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Device/Monitor/DSIPanel/CBiosHX8392A.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Device/Monitor/DSIPanel/CBiosHX8392A.c:* Copyright 2013-2017 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
cbios/Device/Monitor/DSIPanel/CBiosNT35595.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Device/Monitor/DSIPanel/CBiosNT35595.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Device/Monitor/DSIPanel/CBiosNT35595.c:* Copyright 2013-2017 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
cbios/Device/Monitor/DSIPanel/CBiosR63319.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Device/Monitor/DSIPanel/CBiosR63319.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Device/Monitor/DSIPanel/CBiosR63319.c:* Copyright 2013-2017 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
cbios/Device/Monitor/DSIPanel/CBiosR63417.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Device/Monitor/DSIPanel/CBiosR63417.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Device/Monitor/DSIPanel/CBiosR63417.c:* Copyright 2013-2017 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
cbios/Device/Monitor/EDPPanel/CBiosITN156.c:* Copyright 2015 VIA Technologies, Inc. All Rights Reserved.
cbios/Device/Monitor/EDPPanel/CBiosITN156.c:* Copyright 2015 S3 Graphics, Inc. All Rights Reserved.
cbios/Device/Monitor/EDPPanel/CBiosITN156.c:* Copyright 2015 VIA Alliance(Zhaoxin)Semiconductor Co., Ltd. All Rights Reserved.
cbios/Device/Port/CBiosCRT.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Device/Port/CBiosCRT.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Device/Port/CBiosCRT.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Device/Port/CBiosCRT.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Device/Port/CBiosDP.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Device/Port/CBiosDP.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Device/Port/CBiosDP.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Device/Port/CBiosDP.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Device/Port/CBiosDSI.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Device/Port/CBiosDSI.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Device/Port/CBiosDSI.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Device/Port/CBiosDSI.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Device/Port/CBiosDSI.h:* Copyright 2013-2017 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
cbios/Device/Port/CBiosDVO.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Device/Port/CBiosDVO.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Device/Port/CBiosDVO.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Device/Port/CBiosDVO.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Display/CBiosDisplayManager.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Display/CBiosDisplayManager.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Display/CBiosDisplayManager.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Display/CBiosDisplayManager.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Display/CBiosMode.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Display/CBiosMode.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Display/CBiosMode.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Display/CBiosMode.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Display/CBiosPathManager.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Display/CBiosPathManager.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Display/CBiosPathManager.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Display/CBiosPathManager.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/CBiosChipFunc.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/CBiosChipFunc.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/CBiosChipFunc.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/CBiosChipFunc.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/CBiosHwShare.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/CBiosHwShare.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/CBiosHwShare.h:* Copyright 2013-2017 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
cbios/Hw/CBiosPWM.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/CBiosPWM.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/CBiosPWM.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/CBiosPWM.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/Chx001/CBiosVCP_chx.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/Chx001/CBiosVCP_chx.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/Chx001/CBiosVCP_chx.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/Chx001/CBiosVCP_chx.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/Chx001/CBios_chx.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/Chx001/CBios_chx.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/Chx001/CBios_chx.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/Chx001/CBios_chx.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosDIU_CRT.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosDIU_CRT.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosDIU_CRT.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosDIU_CRT.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosDIU_CSC.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosDIU_CSC.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosDIU_CSC.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosDIU_CSC.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosDIU_DP.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosDIU_DP.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosDIU_DP.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosDIU_DP.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosDIU_DVO.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosDIU_DVO.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosDIU_DVO.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosDIU_DVO.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosDIU_HDAC.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosDIU_HDAC.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosDIU_HDAC.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosDIU_HDAC.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosDIU_HDCP.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosDIU_HDCP.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosDIU_HDCP.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosDIU_HDCP.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosDIU_HDMI.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosDIU_HDMI.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosDIU_HDMI.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosDIU_HDMI.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosDIU_HDTV.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosDIU_HDTV.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosDIU_HDTV.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosDIU_HDTV.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosIGA_Timing.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosIGA_Timing.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosIGA_Timing.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosIGA_Timing.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosPHY_DP.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosPHY_DP.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosPHY_DP.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosPHY_DP.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosScaler.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosScaler.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosScaler.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/HwBlock/CBiosScaler.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/HwCallback/CBiosCallbacksHw.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/HwCallback/CBiosCallbacksHw.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/HwCallback/CBiosCallbacksHw.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/HwCallback/CBiosCallbacksHw.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/HwInit/CBiosInitHw.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/HwInit/CBiosInitHw.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/HwInit/CBiosInitHw.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/HwInit/CBiosInitHw.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/HwInterface/CBiosHwInterface.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/HwInterface/CBiosHwInterface.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/HwInterface/CBiosHwInterface.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/HwInterface/CBiosHwInterface.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/HwUtil/CBiosI2C.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/HwUtil/CBiosI2C.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/HwUtil/CBiosI2C.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/HwUtil/CBiosI2C.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/HwUtil/CBiosUtilHw.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/HwUtil/CBiosUtilHw.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/HwUtil/CBiosUtilHw.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/HwUtil/CBiosUtilHw.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/Interrupt/CBiosISR.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/Interrupt/CBiosISR.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/Register/BIU_CR_C_BUS_registers.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/Register/BIU_CR_C_BUS_registers.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/Register/BIU_HDA_registers.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/Register/BIU_HDA_registers.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/Register/BIU_MM_registers.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/Register/BIU_MM_registers.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/Register/BIU_SBI_registers.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/Register/BIU_SBI_registers.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/Register/BIU_TSR_registers.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/Register/BIU_TSR_registers.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/Register/BIU_VCP_registers.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/Register/BIU_VCP_registers.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/Register/DIU_CR_registers.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/Register/DIU_CR_registers.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/Register/DIU_MM_registers.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/Register/DIU_MM_registers.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/Register/DIU_SR_registers.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/Register/DIU_SR_registers.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/Register/DIU_vga_registers.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/Register/DIU_vga_registers.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/Register/Monitor/CBiosDPCDRegister.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/Register/Monitor/CBiosDPCDRegister.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/Register/pmu_registers.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/Register/pmu_registers.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Hw/Register/pmu_registers_e2k.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Hw/Register/pmu_registers_e2k.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Init/CBiosInit.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Init/CBiosInit.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Interface/CBios.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Interface/CBios.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Util/CBiosEDID.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Util/CBiosEDID.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Util/CBiosEDID.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Util/CBiosEDID.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Util/CBiosGPIO.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Util/CBiosGPIO.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
cbios/Util/CBiosUtil.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
cbios/Util/CBiosUtil.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
core_errno.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
core_errno.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
core_errno.h:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
core_module.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
core_module.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
core_module.c:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
kernel_import.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
kernel_import.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
kernel_import.h:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
kernel_interface.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
kernel_interface.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
kernel_interface.h:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
os_interface.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
os_interface.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
os_interface.c:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
os_interface.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
os_interface.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
os_interface.h:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
os_shared.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
os_shared.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
os_shared.c:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
zx.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
zx.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
zx.h:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
zxg.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
zxg.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
zxg.c:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
zxgfx_trace.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
zxgfx_trace.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
zxgfx_trace.h:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
zxgfx_trace_events.c:/** Copyright (c) 2010-2012, S3 Graphics, Inc.
zxgfx_trace_events.c:* Copyright (c) 2010-2012, S3 Graphics Co., Ltd.
zx_backlight.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
zx_backlight.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
zx_backlight.c:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
zx_backlight.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
zx_backlight.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
zx_backlight.h:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
zx_cbios.c: * ** Copyright (c) 2001-2003, S3 Graphics, Inc.
zx_cbios.c: * ** Copyright (c) 2004-2008, S3 Graphics Co., Ltd.
zx_cbios.h: * ** Copyright (c) 2001-2003, S3 Graphics, Inc.
zx_cbios.h: * ** Copyright (c) 2004-2008, S3 Graphics Co., Ltd.
zx_cec.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
zx_cec.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
zx_cec.c:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
zx_cec.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
zx_cec.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
zx_cec.h:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
zx_debugfs.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
zx_debugfs.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
zx_debugfs.c:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
zx_debugfs.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
zx_debugfs.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
zx_debugfs.h:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
zx_def.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
zx_def.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
zx_def.h:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
zx_driver.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
zx_driver.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
zx_driver.c:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
zx_driver.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
zx_driver.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
zx_driver.h:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
zx_ioctl.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
zx_ioctl.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
zx_ioctl.c:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
zx_ioctl.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
zx_ioctl.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
zx_ioctl.h:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
zx_keymap.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
zx_keymap.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
zx_keymap.c:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
zx_led.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
zx_led.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
zx_led.c:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
zx_led.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
zx_led.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
zx_led.h:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
zx_pcie.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
zx_pcie.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
zx_pcie.c:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
zx_sync.c:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
zx_sync.c:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
zx_sync.c:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
zx_sync.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
zx_sync.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
zx_sync.h:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.
zx_types.h:* Copyright 1998-2014 VIA Technologies, Inc. All Rights Reserved.
zx_types.h:* Copyright 2001-2014 S3 Graphics, Inc. All Rights Reserved.
zx_types.h:* Copyright 2013-2014 Shanghai Zhaoxin Semiconductor Co., Ltd. All Rights Reserved.


This should cut off all speculations that the new driver is not based on S3G's IP. As you can see, it is still the same code base, just refactored and enhanced.

[Loeschzwerg]

Zhaoxin does have a new website and they now provide driver downloads for different OS (Win10, 7, XP(!!!), UOS, Deepin, Ubunut): http://www.zhaoxin.com/qdxz.aspx?nid=31&typeid=153

From the most recent Windows 10 driver:

2、CHANGE HISTORY

10.17.19.0024-37.00.24(03/26/2020)
1. Remove 4k mode for customer request
2. Test with Win10 20H1 RC OS

10.17.19.0023-37.00.23(03/13/2020)
1. Run High Definition Audio Device (onboard HDMI/DP) > Audio Codec - Wave Test - Certification :Fail
2. HDMI 4k@60HZ no display

10.17.19.0021-37.00.21(11/13/2019)
1. Customize Option setting between system BIOS and UMA GFX for OEM feature

Note:VBIOS D8.F2.00.21必须搭配GOP V09及之后的版本 , windows driver 37.00.21
及之后的版本, linux driver 20.00.20 及之后的版本

10.17.19.0020-37.00.20(09/27/2019)
1. Add Brightness control by GFX PWM
2. HDMI Electrical Testing Issue in 297MHz(4K x 2K @30)
3. Paint 3D occurs blurred screen when zoom out to below 33%
4. The UUT pops up the error box when running 3DMARK. Line:728 ERROR information

10.17.19.0017-37.00.17(06/20/2019)
1. Test with Win10 19H1

10.17.19.0015-37.00.15(04/29/2019)
1. Eclk=300MHz

10.17.19.0007-37.00.07(11/30/2018)
1. RC1 release with RS5
2. Eclk=350MHz

10.17.19.0005-37.00.05
1. RC release with RS4

10.17.19.0004-37.00.04
1. Beta release with RS4
2. Change GPU name
3. Update GPU clk

10.17.19.0002-37.00.02
1. Alpha release
2. Add 4k@60Hz support

10.17.19.0000-37.00.00
1. Initial release

[Tralalak]

ASRock Centerm S7-6780A TBD with ZHAOXIN KaiXian KX-U6780A
source: https://ranker.sisoftware.co.uk/show_system.php?q=cea598ab93ab92a595b3d4e9c4f5d3a19cad8be2dfefc9a19cac8af2cfffd9 bcd9e4d4f281bc84&l=en

[Benutzername]

Zhaoxin does have a new website and they now provide driver downloads for different OS (Win10, 7, XP(!!!), UOS, Deepin, Ubunut): http://www.zhaoxin.com/qdxz.aspx?nid=31&typeid=153

From the most recent Windows 10 driver:

Not too surpsrising they still offer WinXP drivers. In industry there are lots of machines for CNC machines and stuff still running XP and its variants. also governement offices are also still using XP. Global marketshare is put at 1,5%-2% what a quick search for statistics showed. I could not find stats for China only quickly. I presume the marketshare there is still much higher.


And currently the Zhaxin PCUs are only good for an office PC or industrial applications and such in my humble opinion.



edit: I just remembered that at a friend's workplace they have several CNC machines controlled with WinNT machines. As long as tehy are not connected to a netwrok that is not much of a problem.

[y33H@]

What would be the results with DDR4-2400 @ 18-15-17-39 in dual channel connection?My HX002EA board should arrive tomorrow, it's a KX-U6780A with dual channel for regular U-DIMMs, no SO-DIMMs.

[Benutzername]

My HX002EA board should arrive tomorrow, it's a KX-U6780A with dual channel for regular U-DIMMs, no SO-DIMMs.

I hope with an UEFI with more options to work on RAM timings.

[rasti666]

Tralalak - what is ASRock Centerm S7-6780A TBD. What product is this? Is this again just a preliminary project? My VIA EPIA-M920 is getting old, and here still nothing new on the market ;). Where is this product from ASRock? Where is he from - does anyone know? ASRock knows that something like this is to produce? Because he doesn't write anything about it ;)

[Screemer]

tralalak ist aus tschechien, wenn ich mich richtig erinnere.

[Tralalak]

Tralalak is from Slovakia.

Asrock has always made consumer super low end products with VIA processors so you will be surprised what they come up with Zhaoxin processors.

[rasti666]

The only motherboard I have from ASRock is https://www.asrock.com/mb/VIA/PV530/index.asp. I don't know other motherboards with VIA processors from the ASRock market to buy. I am waiting. ASRock is my favorite producer. Maybe as the only one - it will actually give something to the market again ;)

[Tralalak]

And ITX version:
https://www.asrock.com/MB/VIA/PV530-ITX/index.asp
https://www.asrock.com/mb/VIA/PV530A-ITX/index.asp

Anyway yesterday was video on zhaoxin website and there was graphs

兆芯官方KX-7000是6nm工艺主频4.XG = Zhaoxin official KX-7000 is 6nm process frequency 4.XG

I didn't manage to make a print screen. :-(
They delete it!

[Tralalak]

Zhaoxin to Design Discrete GPUs
source: https://www.techpowerup.com/269657/zhaoxin-to-design-discrete-gpus

[Loeschzwerg]

Zhaoxin to Design Discrete GPUs
source: https://www.techpowerup.com/269657/zhaoxin-to-design-discrete-gpus

Nothing new here, since you've found that piece of information quite some time ago :)
https://www.forum-3dcenter.org/vbulletin/showthread.php?p=12303913#post12303913

[scx]

Anyway yesterday was video on zhaoxin website and there was graphs

兆芯官方KX-7000是6nm工艺主频4.XG = Zhaoxin official KX-7000 is 6nm process frequency 4.XG

I didn't manage to make a print screen. :-(
They delete it!

Are you talking about this video?
https://www.bilibili.com/video/BV1H5411e7mw
https://space.bilibili.com/98717142

In fact, there was something about "4.xGHz" and "7nm/6nm" process.
https://i.ibb.co/tYk19Xs/zhaoxin-2020-07-08-06.png

It is an extended version of the video published on the Zhaoxin website (18:10 vs 13:26).
http://www.zhaoxin.com/spzx_view.aspx?nid=2&typeid=174&id=1138
https://v.qq.com/x/page/s3114mv4nf7.html
https://v.qq.com/vplus/091e557d8eda0a8667d5bd9617482a49

Anyway, it is based on the PDF that was already available in March, with some minor changes.
http://www.gdinsa.org/uploadfile/2020/0326/20200326055237805.pdf

All pages from this video presentation:
https://ibb.co/album/Gvkrk2?sort=name_asc

https://i.ibb.co/x1BX5vs/zhaoxin-2020-07-08-03.png (https://ibb.co/5rSGcD9)
https://i.ibb.co/L1j7NtW/zhaoxin-2020-07-08-04.png (https://ibb.co/KW4vKqk)
https://i.ibb.co/PrYVghW/zhaoxin-2020-07-08-05.png (https://ibb.co/vcBMkPd)
https://i.ibb.co/SVWZzrn/zhaoxin-2020-07-08-06.png (https://ibb.co/TPpNsL8)
https://i.ibb.co/rm7dmBR/zhaoxin-2020-07-08-07.png (https://ibb.co/NsVTs0d)
https://i.ibb.co/QdVWxgQ/zhaoxin-2020-07-08-08.png (https://ibb.co/TLGQNf2)
https://i.ibb.co/pX57psq/zhaoxin-2020-07-08-09.png (https://ibb.co/ryNjJL9)
https://i.ibb.co/c68x0tF/zhaoxin-2020-07-08-10.png (https://ibb.co/7npbPVW)
https://i.ibb.co/hXdXF0T/zhaoxin-2020-07-08-11.png (https://ibb.co/7gpgth0)
https://i.ibb.co/S6hmSPv/zhaoxin-2020-07-08-12.png (https://ibb.co/WHbzC6K)
https://i.ibb.co/hmzyWyS/zhaoxin-2020-07-08-13.png (https://ibb.co/nLJbcbF)
https://i.ibb.co/5MwM334/zhaoxin-2020-07-08-14.png (https://ibb.co/1vjvNNQ)
https://i.ibb.co/94xt7tV/zhaoxin-2020-07-08-15.png (https://ibb.co/B6hN9Nr)
https://i.ibb.co/Qmghc2r/zhaoxin-2020-07-08-16.png (https://ibb.co/gg2Q75z)
https://i.ibb.co/6bdTKh0/zhaoxin-2020-07-08-17.png (https://ibb.co/Gkh8ZwQ)

[Tralalak]

Are you talking about this video?
https://www.bilibili.com/video/BV1H5411e7mw
https://space.bilibili.com/98717142

In fact, there was something about "4.xGHz" and "7nm/6nm" process.
https://i.ibb.co/tYk19Xs/zhaoxin-2020-07-08-06.png

It is an extended version of the video published on the Zhaoxin website (18:10 vs 13:26).
http://www.zhaoxin.com/spzx_view.aspx?nid=2&typeid=174&id=1138
https://v.qq.com/x/page/s3114mv4nf7.html
https://v.qq.com/vplus/091e557d8eda0a8667d5bd9617482a49

Anyway, it is based on the PDF that was already available in March, with some minor changes.
http://www.gdinsa.org/uploadfile/2020/0326/20200326055237805.pdf

All pages from this video presentation:
https://ibb.co/album/Gvkrk2?sort=name_asc


...

https://i.ibb.co/SVWZzrn/zhaoxin-2020-07-08-06.png (https://ibb.co/TPpNsL8)

...


:up:

[y33H@]

Another review with the Cjoyin C1888 and KX-6780A:

https://www.hardwareluxx.de/index.php/artikel/hardware/prozessoren/53677-ein-exot-aus-china-mit-acht-kernen-der-zhaoxin-kx-u6780a-im-test.html

[Tralalak]

Hi guys
Have you had some problems with games like artifacts and etc. Isn't that a GPU problem? And not the CPU KX-U6780A?

https://twitter.com/IanCutress/status/1288212727245213696

P.S. In twenty years, I've never had problems with artifacts with VIA CPUs. It has always been a problem of GPU resp. GPU drivers...

[Loeschzwerg]

Looks like a driver issue to me. I didn't have any troubles so far, but I'm currently using an old GTX745 (upgrade to a GTX 1660S is planned soon).

[y33H@]

No trouble here but I did't bench RDR2 ^^

[Tralalak]

KX-U6780A is no longer the last. :weg:

source: https://www.hardwareluxx.de/index.php/artikel/hardware/prozessoren/53744-lakefield-im-test-core-i5-l16g7-mit-fuenf-kernen-muss-zeigen-was-er-kann.html?start=1


I know 5W vs 70W. But a graph is a graph. :cool:

[y33H@]

Haha, take a single core Bonnell-based Atom ^^

[Benutzername]

KX-U6780A is no longer the last. :weg:

source: https://www.hardwareluxx.de/index.php/artikel/hardware/prozessoren/53744-lakefield-im-test-core-i5-l16g7-mit-fuenf-kernen-muss-zeigen-was-er-kann.html?start=1


I know 5W vs 70W. But a graph is a graph. :cool:


As expected from the one big and four little cores of the i5-L16G7. I wonder what the Chip could do at 70 W? Smoking the Zhaoxin, but by how much?

[Tralalak]

De Zhaoxin KX-U6780A gereviewd
source: https://tweakers.net/reviews/8028/5/de-zhaoxin-kx-u6780a-gereviewd-de-opmars-van-chinese-processors-testresultaten.html


___________



DuxFYDR6Ge4

[Tralalak]

BEqSHwk93lE

[y33H@]

Haha, I had the same board ^^

71158

[Tralalak]

:up:

P.S.dual channel memory configuration ?

[y33H@]

Yes.

[Tralalak]

Have you any benchmarks results between first mainboard (single channel) and second mainboard (dual channel) or with better - faster memory banks ?

[Tralalak]

New Geekbench 5 results
CentaurHauls 2000 MHz (8 cores) vs ZHAOXIN KaiSheng KH-37800D 2700 MHz (16 cores)
CentaurHauls Family 6 Model 71 Stepping 2 vs 2x CentaurHauls Family 7 Model 11 Stepping 14
8x New BIG Cores vs 16x small Cores (2x 8 small Cores)

Multi-Core Score 3330 (+8,82%) vs 3060 (91,89%)
Single-Core Score 482 (+33,15%) vs 362 (75,10%)

source: https://browser.geekbench.com/v5/cpu/compare/3358994?baseline=3361909

P.S. The new KaiSheng KH-40000 series will have up to 32 of these new BIG cores while in dual socket up to 64 cores... :weg:

[y33H@]

Have you any benchmarks results between first mainboard (single channel) and second mainboard (dual channel) or with better - faster memory banks ?The first actually was Dualchannel, as it was already achieving more than 30 GB/s with DDR4-2666. I realized that during testing the second board, so I stopped doing additional benchmarks.

[Tralalak]

CentaurHauls CPU @2200MHz (8C 2.2GHz, 8x 256kB L2, 16MB L3)


Higher than 93.93% ranked results
link: https://ranker.sisoftware.co.uk/show_run.php?q=c2ffcdfcdabbdae7d6e5d6e0d7e3c5b78aba9cf99ca191b7c4f9c1&l=en


Higher than 99.87% ranked results
link: https://ranker.sisoftware.co.uk/show_run.php?q=c2ffcdffd9b8d9e4d5e5d1e2d2e2d0f684b989afcaaf92a284f7caf2&l=en


Higher than 93.06% ranked results
link: https://ranker.sisoftware.co.uk/show_run.php?q=c2ffc8ffd9b8d9e4d3e5d1e9dff98bb686a0c5a09dad8bf8c5fd&l=en


P.S. Qualification: Excellent Performance :weg:

[Tralalak]

Notebook ZHAOXIN KaiXian KX-6640MA 2.2GH+ (4C/4T, 2.6GHz Turbo) processor benchmarked
source: https://tieba.baidu.com/p/6997500663

[Tralalak]

Notebook user review with Zhaoxin KaiXian KX-6640MA 2.2+GHz (4C/4T, up to 2.6GHz Turbo)

rdAMO-imQb8

[Tralalak]

SIV - System Information Viewer Version 5.52 released on 14-Oct-2020

Added CentaurHauls CHA memory controller speed + timing information and Dual Socket system support.


http://rh-software.com/siv_rvc.png

source: http://rh-software.com/

[konkretor]

https://www.digitimes.com.tw/tech/dt/n/shwnws.asp?CnlID=1&id=0000596867_PLH9NBRA5EIV268D2B07R&cf=A61


VIA sells x86 technology to Shanghai Semiconductor, a trade war push in the background.

Yuyuan Chen／Hsinchu 2020-10-27 07:44

VIA sells x86 platform technology to Shanghai Zhaoxin.

VIA announced today that it is holding a press conference to provide information about the transaction, which will be hosted by VIA Chairman Wenqi Chen. VIA's 100% owned subsidiaries, VIABASE and VIATECH, have sold certain intellectual property rights (IPR) related to their chipsets to Shanghai Zhaoxin, a company in which VIA holds an indirect 14.75% stake, for a total transaction amount of approximately US$138 million. The total value of the transaction was US$257 million for a profit of US$197 million, or about NT$5.655 billion.

The total transaction amounted to US$257 million for a gain of US$197 million, or approximately NT$5.655 billion. VIA emphasized that the transaction is an asset revitalization, which will increase revenue and will not affect future business development, and that VIA and its subsidiaries will be able to continue their existing businesses after completion of the transaction.

As to whether the transaction is related to the de-beautification of China, Mr. Chen said that the market has changed a lot and is still under continuous observation. In addition, the Ministry of Economic Affairs (MOEA) has proposed to amend Article 5 of the "Permitted Regulations for Investment or Technical Cooperation in the Mainland" to require prior approval for the transfer of know-how, patent rights, or integrated circuit layout rights in China, in order to prevent the outflow of advanced technologies and plug the loopholes, and to make people wonder whether VIA will conduct the transaction before the amendment is passed. VIA said the deal does not require prior application, but only filing.

VIA is now fully engaged in the AIoT battlefield and can sell its x86 processor, chipset technology, data and other intellectual property rights to Zhaoxin for a good price of NT$5.6 billion, which, according to VIA, will not affect the business of VIA and its subsidiaries and is considered a worthwhile deal. VIA and the Shanghai National AIC Joint Venture received a RMB 5.7 billion investment boost from the Chinese government.

VIA, Shanghai SASAC Joint Venture, Zhaoxin Receives RMB 5.7 Billion Investment Boost

Shanghai Zhaoxin was established in 2013 as a joint venture between VIA and Shanghai Lianhe Investment Company, a subsidiary of Shanghai Municipal Assets Supervision and Administration Commission (SASAC), with a share capital of US$250 million. Zhaoxin has focused on x86 processor development with technology licenses from VIA, but has yet to make a profit.

In 2006, China published the medium- and long-term science and technology development plan (2006-2020), which proposed the implementation of 16 special projects, with a total investment budget of more than 100 billion yuan; the "nuclear high base" plan was the first of 10 special projects, responsible for the development of high-end chips, core components and basic software and other key core technologies. 2014, the one-year-old Zhaoxin, on the VIA received the "Nuclear High-Tech Foundation" project 01, and worked with the Shanghai Institute of Advanced Studies, Chinese Academy of Sciences, Lenovo Group and Shanghai Hua Li Microelectronics to develop desktop computer (DT) software and hardware research and applications.

At the time, VIA only stated that it had IC design expertise and marketing teams and was interested in seeking a China partner to secure funding and expand into the China market, while Shanghai Lianhe Investment, which has deep roots in the China market and is also interested in investing in the IC industry, expressed interest in the partnership but did not specify the details. Although VIA's processor platforms are far less powerful than those of Supermicro and Intel, they still have the basic technology development capabilities that were expected to help boost China's PC processor development capabilities.

China-U.S. trade war resumes Mega Core fears to follow Haiguang by the U.S. lockout

It's worth noting that the cross-licensing agreement between VIA and Intel has long been a source of confusion, and Intel has agreed to sub-license MHz, with no clear explanation from the three parties. Extends the license for VIA x86 and retains the PCI-Express bus as a key interface for at least six years. VIA's joint venture with the Chinese government to set up Zhaoxin was probably on the fringes of the law at the time, but it was also aimed at Intel, which at the time relied heavily on the Chinese market and was worried about the resurgence of monopoly concerns, and did not want to pressurize Shanghai Lianhe Investment, which has a Chinese background.

However, with the start of the US-China trade war, Tianjin Haiguang, which has also acquired a supercomputers x86 license, mainly for servers, has been clamped down by the US government in 2019 and is now on the list of entities, making the next step in Megachip's x86-based processor platform development a major concern.

VIA's stock king is no longer in his prime, and his focus once shifted to China.

Founded in 1992, VIA entered the x86 processor and chipset design arena in 1999 and 2000 with the acquisition of National Semiconductor's Cyrix X86 processor division, IDT Processor Corporation and S3 Graphics. However, the third-party chipset business, which supplied both Supermicro and Intel, brought VIA a temporary boom around the year 2000, with profits jumping dramatically and the company's share price climbing to the top of the market at more than NT$600, giving it a market value of more than NT$100 billion.

However, the good times did not last forever, as VIA and Intel were involved in litigation in 2001, and the processor business was also blocked, the situation took a sharp turn for the worse, with the stock price plummeting. In addition, VIA has been supplying Southbridge chips for the Dragon Core series of MIPS-based processors, which were previously developed and manufactured in China.

VIA launches AIoT transformation plan to reap benefits from outside the industry

VIA finally turned profitable since 2015, but the industry did not get rid of its losses. The key to the turnaround came from the recognition of CDMA licenses granted by VIA to MediaTek, the sale of CDMA-related technology to Intel by its VIA Telecom subsidiary, and the settlement of a lawsuit with Xiangshang in 2017 for tens of millions of dollars. VIA's product strategy has been revised over the years, from semiconductor design to embedded platforms.

After years of adjustment, VIA has revised its product strategy from semiconductor design to embedded platform, and then to the latest AI solutions, focusing its core technologies on AI, computer vision, deep learning and machine learning.

It is worth noting that VIA's subsidiary, VIA Technologies, which is listed on the Emerging Stock Exchange, will soon be listed on the Taiwan Stock Exchange (TSE). VIA's main business focuses on high-speed transmission interface ICs. Benefiting from the increased penetration of USB Type-C and USB3.1 terminal applications, VIA's first-half revenue in 2020 will reach $933 million, an annual increase of 35%, and net profit of $156 million, an annual increase of 49%.

Translated with www.DeepL.com/Translator (free version)

[Tralalak]

VIA / Centaur is still going ahead with CNS, they have a conference presentation tomorrow on it

[y33H@]

It took me quite some time, but that's the actual source:

https://emops.twse.com.tw/server-java/t05st01_e

1.Date of occurrence of the event:2020/10/26
2.Company name:VIA Technologeis, Inc.
3.Relationship to the Company (please enter ”head office” or
”subsidiaries”):VIA Technologeis, Inc.and its subsidia
4.Reciprocal shareholding ratios:N/A
5.Cause of occurrence:
The company held a major information explanatory press conference on the
TWSE at 9:00 pm on 2020/10/26.
The statement regarding the resolution of the company's board of directors
to dispose of intangible assets is as follows:
The company and its 100% owned subsidiaries VIABASE CO., LTD. and VIATECH
CO., LTD. sell some chipset products related intellectual property rights
(excluding patent rights) to Shanghai Zhaoxin Semiconductor Co., Ltd., in
which VIA has indirect shareholding of 14.75% the transaction amount
is respectively VIA US$30,574 thousand、VIABASE US$26,405 thousand and
VIATECH US$81,995 thousand. The total transaction amount is US$ 138,974
thousand.
In addition, the subsidiary VIABASE CO., LTD. sold some cpu products related
intellectual property rights (excluding patent rights) to Shanghai Zhaoxin
Semiconductor Co., Ltd., the total transaction amount is US$118,418thousand.
The estimated consolidated after-tax profit of the disposal are approx.
US$106,628 thousand and US$90,856 thousand respectively.
The company has a significant influence on counterparty. Therefore, the
unrealized interest US$18,449thous and US$15,720 thousandis expected to be
eliminated based on the shareholding ratio 14.75%.

[Tralalak]

15W 16nm KX-6000G Series with new DX12 iGPU @ Q4/2021
6nm KX-7000 Series base 3.0GHz (~4.1GHz Turbo) with HP iGPU, DDR5, PCIe 4.0 @ Q4/2022

source: https://tieba.baidu.com/p/7051349031

[RavenTS]

Das ist ja schon noch ein ganzes Stück hin ... aber die Turbo-Angaben sind dann schon so fix vorhersehbar.?!:freak:

[Tralalak]

Zhaoxin processors @ KINO mainboard in Germany

KINO-KX-U6580-R10
Mini-ITX Board Zhaoxin KX-U6580 2.5GHz/VGA/DVI


KINO-KX-U6780A-R10
Mini-ITX Board Zhaoxin KX-U6580A 2.7GHz/VGA/DVI

source: https://www.icp-deutschland.de/?cl=search&searchparam=%2A&filterTechnologien=Zhaoxin&navigation=true
source: https://files.icp-deutschland.de/produkte/KC002667/web/icp/KINO-KX-datasheet-20201023.pdf

[Loeschzwerg]

Ahhh, that's interesting. So Zhaoxin SoCs aren't limited to the Asian market :)

[Benutzername]

Ahhh, that's interesting. So Zhaoxin SoCs aren't limited to the Asian market :)


If an importer finds it worthwhile to start improting the boards the Zhaxin has a distributor. Makes me wonder how many they expect to sell? Except for tinkerers and collectors of obscure and exotic hardware I do not really see much of a market.

[Loeschzwerg]

You could ask the same "room for a market" question for nearly every embedded system / board. If the SoC is cheap enough and fits the task, why not. There are some really really strange products out there ^^ (e.g. even with Vortex86, ZFx86, RDC stuff)

The main thing however is, that this board meets compliance regulations (CE, FCC, RoHS and so on), which means that it can be officially sold in Europe / US. Whereas the Cjoyin C1888 is only for Asian market. It's not simply a distributor thing.

[Tralalak]

Hi guys. What is the result of your KX-U6780A in Cinebench R23?
VIA QuadCore C4650 2.0GHz

Single-Core: 209
Multi-Core: 733

https://i.imgur.com/sLc0tvJ.png
https://i.imgur.com/kOCiF2X.png

[Tralalak]

Geekbench 5 results Linux
ZHAOXIN KaiSheng KH-37800 @ 2.70 GHz vs Centaur @ 2.20 GHz
2 processors, 16 cores vs 1 processor, 8 cores

Single-Core Score: 348 (66,41 %) vs 524 (+50,57 %)
Multi-Core Score: 3 747 (96,85 %) vs 3 869 (+3,26 %)

source: https://browser.geekbench.com/v5/cpu/compare/341928?baseline=4697254

[Tralalak]

-yRvyQ8wCYs

[Tralalak]

"Fingers crossed for the Centaur CNS. We should be getting one soon."
-Ian Cutress, Nov 30th | Senior CPU Editor, http://AnandTech.com | TechTechPotato

source: https://twitter.com/IanCutress/status/1333618256851165189

[y33H@]

Well, let's hope he'll do a review - he never published the U6880A ^^

[Tralalak]

May be he want compare Zhaoxin KX-U6780A processor (CentaurHauls) with new Centaur CHA processor (CentaurHauls - CNS) in one review.

[y33H@]

He may do that, but he has the Zhaoxin for a few months now ... without knowing if he will ever get the Centaur, too.

[Tralalak]

I can assure you he knew that he would get it. They had promised it. :whisper:

[Tralalak]

I remember to my ZBOX nano VD01 mini-PC (https://www.zotac.com/ae/product/mini_pcs/vd01)

source: http://www.szrealan.com/show-214.html

[Benutzername]

He may do that, but he has the Zhaoxin for a few months now ... without knowing if he will ever get the Centaur, too.

May be he want compare Zhaoxin KX-U6780A processor (CentaurHauls) with new Centaur CHA processor (CentaurHauls - CNS) in one review.

IIRC in his Q&A video series on his youtube channel Dr. cutress made an offhand comment about having to write the Zhaoxin board article still, having done tests on it. Well if he is going to write one bigger article for both boards that would be nice of course. But I am afraid it will be delayed by other matters again and again.



That video that Tralalak linked was interesting. (y) Although I do not understand why they stick to x86 for the machine learning/AI stuff target now. Those machines run in specialised computing centres with special software anyway. So they could use any instruction set really. Others use ARM with their own ML/AI extensions as far as i know.

[Tralalak]

Rumours: ZHAOXIN Kaisheng KH-40000 Series Processor

Single-core performance increased by 60%, multi-core performance increased five times.
Multi-channel connection, 16 memory channels, 128-stream PCIE
LGA plug-in package

image: http://tieba.baidu.com/photo/p?kw=%E5%85%86%E8%8A%AF&ie=utf-8&flux=1&tid=7165060551&pic_id=3f8e02fb513d2697cde6e21842fbb2fb4316d826&pn=1&fp=2&see_lz=1

source: https://tieba.baidu.com/p/7165060551
source: https://www.bilibili.com/video/BV14X4y1T7NQ?from=search&seid=11337936780182631934

[rasti666]

Hello.

I just purchased a KINO-KX-U6580-R10. When I get it, I will share the test results with you ;-)

Greetings.

P.S. My VIA EPIA M920-20Q has a successor after many years ;-).

[Loeschzwerg]

Nice :) How much does it cost?

If you need drivers for the IGP, let me know.

[rasti666]

I paid 423 dollars. It is a lot of money, but for a true VIA fan - not a problem;).

Loeschzwerg - thank you, I'll let you know when I need the drivers.

[rasti666]

Hello,

I am just building my new computer, but I did a quick test in CineBench R20,

- System - Windows 7 Professional 64-bit

CINEMA-KX-U6580-R10:
ZHAOXIN KaiXian KX-U6580 (8x2.5 GHz)

GOODRAM 16GB (2x8 GB RAM DDR4 2666 MHz)

Single Core - 118
Multiple Core - 920
MP Ratio 7.77x

I will do more thorough tests when I assemble everything ;)

[rasti666]

I did some tests on UBUNTU 20.04.2 LTS.

I have used "PTS" tests - Phoronix Test Suite in the latest version 10.2.1. For comparison, I tested my second computer at the company.

ZHAOXIN KaiXian KX-U6580 - 8 x 2.5 GHZ
GOODRAM 2 x 8GB 2666 MHz

and

AMD FX-4300 - 4 x 3.8 GHz
GOODRAM 1 x 8GB 1333 MHz

1) https://openbenchmarking.org/test/pts/c-ray
(1.1 - v1.2.0)

330.559 seconds (ZHAOXIN)
470.338 seconds (AMD)

ZHAOXIN is faster;)

2) https://openbenchmarking.org/test/pts/compress-7zip
(16.02 - v1.7.1)

13292 MIPS (ZHAOXIN)
10569 MIPS (AMD)

ZHAOXIN is faster again;)

It seems that the KX-U6580 is already a fast processor, beating the old - but good competition;).

[Skysnake]

Single vs dual channel...

[Tralalak]

Zhaoxin KiaXian KX-U6780A 2.7GHz in the Anandtech CPU 2021 Benchmarks Database
source: https://www.anandtech.com/bench/product/2625?vs=2688

[y33H@]

So we finally may see the review ... with just a few months delay :biggrin:

[lobby]

I'd like to know which one of the drivers in the list can make the audio controller work?
I bought a mini pc on taobao, with a zhaoxin 6400ma CPU, and installed opensuse. it works fine. The only problem i met so far is I can't play sound via HDMI. this is because the system can't recognize gpu's built-in audio controller (not the realtek one, sound via 3.5mm is good), so a driver is needed.
Rpm graphics driver pakages are rare and outdated. Debs are newer but in the release note of the most recent ones, sound is still an issue to be tackled.

[lobby]

Nice :) How much does it cost?

If you need drivers for the IGP, let me know.

Do you have the IGP driver for 6400ma?

I need a rpm driver for my opensuse, but a deb works too I guess. I can try to convert it to rpm.

[Loeschzwerg]

Zhaoxin does have drivers for the following Linux distributions:
Kylin V10 (latest driver supports Linux Kernel 4.4.131)
Deepin V15.11 (latest driver supports Linux Kernel 4.15)
Ubuntu 16.04.5 (latest driver supports Linux Kernel 4.15, 4.18 and 4.19)
Ubuntu 18.04 (latest driver supports Linux Kernel 4.15)

All drivers are .deb packages and support the complete ZX-E (KX-6000 series) range of processors. Your best bet is probably the Ubuntu 16.04.5 driver version 21.00.26.

[Tralalak]

SisoftSandra ranker - Processor Multi-Media
● VIA CentaurHauls CPU @2200MHz (8C 2.2GHz, 8x 256kB L2, 16MB L3):
42 313Mpix/s (+278,06%)

vs.

SisoftSandra ranker - Processor Multi-Media
● ZHAOXIN KaiXian KX-U6780A@2.7GHz (8C 2.7GHz, 2x 4MB L2):
11 192Mpix/s (26,45%)

source: https://ranker.sisoftware.co.uk/show_system.php?q=cea598ab93a096a098bed9e4c9f8deac91a187eed3e3c5ad90a086fec3f2d4 b1d4e9d9ff8cb189&l=en
source: https://ranker.sisoftware.co.uk/top_run.php?q=c2ffcee885e8d5ecdeedcbb984b492fbc6f6d0b885b593ebd6e7c1a4c1fcccea99 a494&l=en



SisoftSandra ranker - Processor Image Processing (Normal/Single Precision)
● VIA CentaurHauls CPU @2200MHz (8C 2.2GHz, 8x 256kB L2, 16MB L3):
32 573Mpix/s (+385,08%)

vs.

SisoftSandra ranker - Processor Image Processing (Normal/Single Precision)
● ZHAOXIN KaiXian KX-U6780A@2.7GHz (8C 2.7GHz, 2x 4MB L2):
6715Mpix/s (20,62%)

source: https://ranker.sisoftware.co.uk/show_system.php?q=cea598ab93a096a098bed9e4c9f8deac91a187eed3e3c5ad90a086fec3f2d4 b1d4e9d9ff8cb189&l=en
source: https://ranker.sisoftware.co.uk/top_run.php?q=c2ffc8fbddb0dde0d9ebd8fe8cb181a7cef3c3e58db080a6dee3d2f491f4c9f9df ac91a9&l=en


EDIT:


MLPerf Inference v1.0: 2000 Suite Results, New Power Measurements

Compare that to Centaur, which has provided a handful of data points for its upcoming CHA AI coprocessor.

● Centaur Integrated x86 CPUs (2.0GHz) - Centaur Integrated AI Coprocessor (2.0GHz)
● Centaur Integrated x86 CPUs (2.5GHz) - Centaur Integrated AI Coprocessor (2.5GHz)

source: https://www.anandtech.com/show/16632/mlperf-inference-v10-2000-suite-results-new-power-measurements
source: https://mlcommons.org/en/inference-edge-10/

[lobby]

Zhaoxin does have drivers for the following Linux distributions:
Kylin V10 (latest driver supports Linux Kernel 4.4.131)
Deepin V15.11 (latest driver supports Linux Kernel 4.15)
Ubuntu 16.04.5 (latest driver supports Linux Kernel 4.15, 4.18 and 4.19)
Ubuntu 18.04 (latest driver supports Linux Kernel 4.15)

All drivers are .deb packages and support the complete ZX-E (KX-6000 series) range of processors. Your best bet is probably the Ubuntu 16.04.5 driver version 21.00.26.

That was a dead end. I found no driver that supports the built-in audio controller. I converted the debs into rpms with alien, but the installation ended up with some internal errors. I believe it was the post-script after normal installation which should be responsible for this.
Perhaps I should try the original UOS system but it's not as fluent as opensuse and I'm crazily busy these days. I will change the os to windows for now and wait until drivers supporting mainstream linux distributions are released. Maybe there won't be that day but I can only hope for that.

[davidzo]

SisoftSandra ranker - Processor Multi-Media
● VIA CentaurHauls CPU @2200MHz (8C 2.2GHz, 8x 256kB L2, 16MB L3):
42 313Mpix/s (+278,06%)

vs.

SisoftSandra ranker - Processor Multi-Media
● ZHAOXIN KaiXian KX-U6780A@2.7GHz (8C 2.7GHz, 2x 4MB L2):
11 192Mpix/s (26,45%)

source: https://ranker.sisoftware.co.uk/show_system.php?q=cea598ab93a096a098bed9e4c9f8deac91a187eed3e3c5ad90a086fec3f2d4 b1d4e9d9ff8cb189&l=en
source: https://ranker.sisoftware.co.uk/top_run.php?q=c2ffcee885e8d5ecdeedcbb984b492fbc6f6d0b885b593ebd6e7c1a4c1fcccea99 a494&l=en



SisoftSandra ranker - Processor Image Processing (Normal/Single Precision)
● VIA CentaurHauls CPU @2200MHz (8C 2.2GHz, 8x 256kB L2, 16MB L3):
32 573Mpix/s (+385,08%)

vs.

SisoftSandra ranker - Processor Image Processing (Normal/Single Precision)
● ZHAOXIN KaiXian KX-U6780A@2.7GHz (8C 2.7GHz, 2x 4MB L2):
6715Mpix/s (20,62%)

source: https://ranker.sisoftware.co.uk/show_system.php?q=cea598ab93a096a098bed9e4c9f8deac91a187eed3e3c5ad90a086fec3f2d4 b1d4e9d9ff8cb189&l=en
source: https://ranker.sisoftware.co.uk/top_run.php?q=c2ffc8fbddb0dde0d9ebd8fe8cb181a7cef3c3e58db080a6dee3d2f491f4c9f9df ac91a9&l=en


EDIT:



Those are some impressive Floating point Numbers :eek:
It is roughly on par with a Ryzen 1700x 8C 3,6Ghz CPU.

Processor Multi-Media AMD Ryzen 7 1700X 44783Mpix/s
Processor Image Processing AMD Ryzen 7 1700X 29700Mpix/s


https://ranker.sisoftware.co.uk/show_device.php?q=c9a598d994d0f0a2dba1c4aa8abd9dac9bab9bc3e582bf92a385f7cafadcb5 88b99ff7caffd9a19cad8bee8bb686a0d3eede&l=en
If you account for the clockspeed disadvantage, 2,2Gz vs 3,6Gz the FPU should be at least as beefy as any modern Zen Architecture including Zen3. Looks like truly a "big core" Design.

There are some strange numbers for arithmetic though. They still have to iron it out, so lets wait and see if the integer and front and backend gets an equally big upgrade.

[Tralalak]

Because Centaur CHA processor (CNS x86 microarchitecture) is first BIG core design to decode up to five instructions per cycle with SIMD as FMA3 up to AVX512 and 16MB L3 cache.

https://i.imgur.com/UU96AOP.png

[davidzo]

Because Centaur CHA processor (CNS x86 microarchitecture) is first BIG core design to decode up to five instructions per cycle with SIMD as FMA3 up to AVX512 and 16MB L3 cache.

https://i.imgur.com/UU96AOP.png

5-decode? That is a crazy wide frontend for an x86, even wider than Zen3 and tigerlakes 4. Wikichiop sys its only 4 wide: https://en.wikichip.org/wiki/centaur/microarchitectures/cha
But does it not have an op cache?

Btw, the 16mb L3 cache also has the same associativity as AMDs Zen3.
It does look fast, but all that in 16nm doesn't sound very power efficient to me.

[Tralalak]

The Centaur CNS already contains four, like Intel processors from Conroe to Haswell or Zen / Zen 2. The decoders can process four instructions, in some cases five, if any two instructions can be merged.

[y33H@]

The decoders typically process four instructions per cycle, but certain pairs of x86 instructions can be decoded together, yielding a maximum of five instructions in a single cycle.

75218

https://www.centtech.com/wp-content/uploads/2020/10/MPR_Centaur_CHA_2020_10_12.pdf

[Tralalak]

CPU-Z 1.96.1
Improved support of Centaur CHA processor
source: https://www.cpuid.com/softwares/cpu-z.html

[Tralalak]

QNAP TVS-675 with Zhaoxin KX-U6580 2.5GHz
source: https://post.smzdm.com/p/a7ddqppl/

[Tralalak]

QNAP TVS-hx75U with Zhaoxin KX-U6580 2.5GHz
source: https://view.publitas.com/qnap-1/tvs-hx75u_cn_20210121/page/1

[Tralalak]

IEI Launched Zhaoxin KX-6000 Series Processor Mini-ITX - KINO-KX Series


https://www.ieiworld.com/_upload/support/images/newsroom/2021/2021_05_KINO-KX/KINO-KX_700X366_ok.jpg

source: https://www.ieiworld.com/en/news/con_show.php?op=showone&cid=1245
source: https://www.ieiworld.com/en/product/model.php?II=739

[rasti666]

Hello

I bought this motherboard in January this year - KINO-KX-U6580-R10 (IEI). It works very well. It is much faster than previous motherboards with VIA processors. After mounting GTX 1050 TI - you can play almost every game ;).

I changed only a factory "very noisy" fan, on a quiet NOCTUA.
Factory cooling was not suitable for use. It was terribly loud ;)

[Skysnake]

What is the idle Power consumption and price tag? If competitive it could be interesting for a nas solution.

But I thought it has no ECC support right?

[Tralalak]

to rasti666

Hi. You can write me the exact tip of the quiet NOCTUA cooler?

My board is on its way for review Rise of the small cores...

Stay tuned

[rasti666]

To Tralalak

I bought this one:

https://noctua.at/en/nf-a8-pwm

Perfectly enters the default heatsink. You can manage the fan speed (PWM) in the BIOS.

In the Phoronix Test Suite on the UBUNTU 20.04.2 LTS - ZHAOXIN KaiXian KX-U6580 wins with the AMD FX-4300 processor - and this is a success for a Chinese manufacturer ;-)

It is the first motherboard from VIA / ZHAOXIN that is world-class.

The previous motherboards I currently have (EPIA-M900-12LQ and EPIA-M920-20Q) are good, but much weaker than the competition ;)

KX-U6580 even in games, won with my old AMD ;)

[Benutzername]

Hello

I bought this motherboard in January this year - KINO-KX-U6580-R10 (IEI). It works very well. It is much faster than previous motherboards with VIA processors. After mounting GTX 1050 TI - you can play almost every game ;).

How many FPS in PUBG? ;)



I changed only a factory "very noisy" fan, on a quiet NOCTUA.
Factory cooling was not suitable for use. It was terribly loud ;)


This is an industrial mobo with its dozens of universal and not universal serial connections. So nobody cares if the fan is loud. It has to be durable and cool the system in a hot factory or hidden in a cabinet or similar. An 80 mm noctua is certainly the enthusiasts solution, but any silent 80 mm fan should do better than the included one.


Does this board have the intel pattern mounting holes like the other ZhaoXin boards? That would open a plethora of cooling options. An Arctic Freezer 11LP for example is a simple two heatpipe cooler that costs new about as much as a brown Noctua. Or pick up any used heatpipe cooler from ebay or something. Heck why not cool it with water? :biggrin:

[rasti666]

To Benutzername:

This cooling is extremely loud. It is not suitable for use even in industry. I've never heard anything like it. This is a product with an 8-core processor for use in industry and multimedia (multimedia walls and the like).

The hole pattern is similar to Intel - I haven't checked it.

Tralalak - he will check it, we know it ;)

I have previous "industrial" motherboards with VIA processors. This one was the loudest.

[Benutzername]

To Benutzername:

This cooling is extremely loud. It is not suitable for use even in industry. I've never heard anything like it. This is a product with an 8-core processor for use in industry and multimedia (multimedia walls and the like).

The hole pattern is similar to Intel - I haven't checked it.

Tralalak - he will check it, we know it ;)

I have previous "industrial" motherboards with VIA processors. This one was the loudest.


Sounds (ha ha ha, pun not intended) like they picked the cheapest fan they could get. :freak:



How much did your board cost?

[Tralalak]

Sneak Peek


https://i.imgur.com/zY9t5pK.jpg


Rise of the small cores :weg:


https://i.imgur.com/1L2pLXT.png


Stay tuned :popcorn:

[rasti666]

I bought this motherboard in January. I paid 423 dollars.

Nevertheless, I'm happy. 14 USB on the back - I needed it ;). They are finally 4 SATA outputs (old models had only 2). I work on Windows 7 Pro (it's perfect) and Ubuntu 20.04.2 LTS.

You can build a normal PC - and it really works well ;).

Before Via / Zhaoxin issued processors under his own logo, I worked on Cyrix.

In new processors - I see huge progress.

[rasti666]

Tralalak - You have a nicer box. I have such "
Traditional industrial "- gray as for EPIA-M920-20Q ;-). You have a colored picture ;-)

[Benutzername]

I bought this motherboard in January. I paid 423 dollars.


In the current market that is a bargain. ;) These boards certainly offer lots of USB and serial port I/O. Is there any atom or embedded Ryzen board having as much?


Nevertheless, I'm happy. 14 USB on the back - I needed it ;). They are finally 4 SATA outputs (old models had only 2). I work on Windows 7 Pro (it's perfect) and Ubuntu 20.04.2 LTS.

I'd rather describe Win7 as less annoying than Win10, Vista or 8. ;)



You can build a normal PC - and it really works well ;).

Before Via / Zhaoxin issued processors under his own logo, I worked on Cyrix.

In new processors - I see huge progress.

Certainly. The process shrinking and raising the frequency alone helps, but they did some serious work on these cores at Centaur. So that should lead to results. Although the starting point was more in intel Atom territory.


Have you done some linux benches? e.g. the phoronix suite, llcbench, sysbench? And how many bogomips do you get? ;)

[rasti666]

Yes, I did some tests. They are in previous posts:

- System - Windows 7 Professional 64-bit

In CineBench R20 :

Single Core - 118
Multiple Core - 920
MP Ratio 7.77x

- System - UBUNTU 20.04.2 LTS.

In "PTS" tests - Phoronix Test Suite in the latest version 10.2.1. :

1) https://openbenchmarking.org/test/pts/c-ray
(1.1 - v1.2.0)

330.559 seconds

2) https://openbenchmarking.org/test/pts/compress-7zip
(16.02 - v1.7.1)

13292 MIPS

It won with the AMD FX-4300 processor in tests.

And this is really progress ;)

[Tralalak]

Cinebench R20 (Multi Core)

KX-U6780A 2.7GHz (Loeschzwerg): 933 cb (+5,30%)
KX-U6580 2.5GHz (rasti666): 920 cb (+3,84%)
KX-U6780A 2.7GHz (y33H@): 886 cb (100%)


The difference of 47 cb is large for the same boards with KX-U6780A. But defference of 34 cb between KX-U6780A 2.7GHz and KX-U6580 2.5GHz is very interesting...

[rasti666]

I think it's RAM.

I used GOODRAM 2 x 8GB 2666 MHz. They come in DUAL and are 2666 MHz

Loeschzwerg had 1x8GB 2400 MHz - I think

It was also a different motherboard ;)

[Tralalak]

130 EUR Zhaoxin Mini-PC
source: https://post.smzdm.com/p/a85rmr57/
link: http://www.zhaoxin.com/news_view.aspx?nid=2&typeid=254&id=1266

[Skysnake]

Hmmn sounds interesting. What is the performance compared to an E8400 core2duo CPU?

For that price it would be worth a look for my mother

[Loeschzwerg]

Looking at my Cinebench R11.5 results (KX-U6870A with 2 active cores -> 1.32 points), the E8400 should beat the KX-6640MA when it comes to a core by core comparison.

[Tralalak]

Windows 11 with Zhaoxin KX-U6780A processor
source: https://tieba.baidu.com/p/7409047490

[Tralalak]

QNAP TVS-675-8G

https://www.ediscomp.sk/img/products/caa/3923979.jpg

source: https://www.ediscomp.sk/produkt/qnap-tvs-675-8g-8c-zhaoxin-kx-u6580-2-5ghz-8gbram-6xsata-2xgbe-2xusb3-0-2xusb3-1-1xhdmi-2xpcie/1671543/

[Skysnake]

Quite expensive for what you get I would say. Then I woukd prefer a self build system.

[Loeschzwerg]

https://twitter.com/Loeschzwerg_3DC/status/1407956912457592838

Does anyone know more about this dGPU project called 极瑞3000 from Zhaoxin?

https://pbs.twimg.com/media/E4oP8gVWQAIdf0m?format=jpg&name=4096x4096

[Tralalak]

This is not Zhaoxin.

[Loeschzwerg]

OK, thanks for clarification. It's 中船重工716 (China Shipbuilding Industry Corporation 716) then?

Edit:
Alright, I found the origin of that picture => http://www.glenfly.com/index.php?catid=22

Glenfield's smart GPU department is the earliest established and most complete graphics processor design and R&D team in China. The R&D department covers architecture, front-end design, back-end design, system engineering, software and testing. The echelon is complete and has extensive experience in chip mass production.

The main products of the GPU team include: discrete graphics chips, X86-SOC and Arm-SOC display chips, covering the discrete, embedded and mobile markets.

[Tralalak]

Preview Rise of the small cores :weg:
IEI KINO-KX-U6580-R10 Mini-ITX mainboard

ZHAOXIN KX-U6580 2.5GHz 8-core processor with C-960 Graphics Processing Unit with 2D/3D/Video Acceleration on-board SoC and Zhaoxin ZX-200 chipset


https://i.imgur.com/OiHca3G.jpg


https://i.imgur.com/QZBPolB.jpg?1


https://i.imgur.com/S62PNu3.jpg?1

Stay tuned :popcorn:

[Skysnake]

Idle Power price and ecc Support please 😉

[Tralalak]

Idle Power price and ecc Support please 😉
370 EUR (+ VAT)
Idle power will see in review...
I thing this board not supported ECC RAM.

[Tralalak]

First impression - Rise of the small cores

Cinebench R23 CPU (Multi Core)
● VIA QuadCore C4650 2.0GHz (28nm, 4C/4T, 2MB L2 cache) ('https://i.imgur.com/kOCiF2X.png'): 733 pts ('https://i.imgur.com/sLc0tvJ.png') (32,91%)
● Zhaoxin KaiXian KX-U6580 (16nm, 8MB L2 cache) ('https://i.imgur.com/TUBf1vG.png'): 2 227 pts ('https://i.imgur.com/UUaDjC7.png') (+203,82%)


And here is my older game results with VIA QuadCore compare with Zhaoxin OctaCore

World of Tanks enCORE - ULTRA
● VIA QuadCore C4650 2.0GHz (28nm, 4C/4T, 2MB L2 cache) + NVIDIA GeForce GTX1050Ti 4GB OC ('https://i.imgur.com/TH4kb91.jpg'): 5 648 ('https://i.imgur.com/AIGOmaA.png') (50,10%)
● Zhaoxin KaiXian KX-U6580 (16nm, 8MB L2 cache) + NVIDIA GeForce GTX1050Ti 4GB OC ('https://i.imgur.com/N6L9N9h.jpg'): 11 273 ('https://i.imgur.com/0GPvnF6.jpg') (+99,59%)

World of Tanks enCORE - MEDIUM
● VIA QuadCore C4650 2.0GHz (28nm, 4C/4T, 2MB L2 cache) + NVIDIA GeForce GTX1050Ti 4GB OC ('https://i.imgur.com/TH4kb91.jpg'): 8 159 ('https://i.imgur.com/5X5Tz1Y.png') (30,56%)
● Zhaoxin KaiXian KX-U6580 (16nm, 8MB L2 cache) + NVIDIA GeForce GTX1050Ti 4GB OC ('https://i.imgur.com/N6L9N9h.jpg'): 26 702 ('https://i.imgur.com/dGY4Syx.jpg') (+227,27%)

[Tralalak]

Sniper Ghost Warrior Contracts 2 @ 2021

1080p HIGH Graphics Settings = perfectly playable and huge graphics


CGL_GloK9zw

[Sardaukar.nsn]

Schön zu sehen das es läuft. Wird sicher noch spannend in den nächsten Jahren.

[Tralalak]

TJWPCTdnjsE


But anyway.
Are you ready for a marathon of the rise of small cores?

Centaur vs Tremont sorry LuJiaZui vs Tremont

Coming soon :weg:

[davidzo]

Nice, i am exited for it!

There are no good Tremont Reviews around, and this architecture is so important for the future for intels microprocessors. Raptor lake will double the small Cores, so we will have many of thm in our everyday PCs in the next five years.

I am also really exited on the centaur cns, but this is considered a big core already, isn't it?

[Skysnake]

Yeah interesting products. The only drawback is the missing ECC and if course we have to check the price, but definitely something I would be happy to test, since it looks at least right now like a promising product

[Tralalak]

I'm very excited with the Intel Pentium Silver N6000 with four Tremont cores.
It will be a very good duel.
I'm not worried about Alder Lake at all.

But let's not get ahead of ourselves.

The marathon of the rise of small cores is just beginning...

Stay tuned and don't go away.

[Tralalak]

Rise of the small cores


https://i.imgur.com/sQuX4Bh.jpg

[Tralalak]

In the footsteps of Alder Lake - The rise of SMALL cores (https://www.forum-3dcenter.org/vbulletin/showthread.php?p=12738821#post12738821)

[Wörns]

Apparently the editors from heise are reading along here:
QNAP-NAS mit chinesischer CPU auch in Deutschland erhältlich (https://www.heise.de/news/QNAP-NAS-mit-chinesischer-CPU-auch-in-Deutschland-erhaeltlich-6143118.html)
MfG

[Tralalak]

Apparently the editors from heise are reading along here:
QNAP-NAS mit chinesischer CPU auch in Deutschland erhältlich (https://www.heise.de/news/QNAP-NAS-mit-chinesischer-CPU-auch-in-Deutschland-erhaeltlich-6143118.html)
MfG

But apparently weak because here is also review the same Zhaoxin KaiXian KX-U6580 2.5GHz processor. (https://www.forum-3dcenter.org/vbulletin/showthread.php?t=607246) (joke) :idea:

[y33H@]

The QNAP NAS is pretty old, don't know why they cover it now ...

EDIT: ah, it's listed in a single German shop ^^

[Tralalak]

5jJ0VoN_YPc

[Tralalak]

The Weird and Wacky World of VIA, the 3rd player in the “Modern” x86 market
source: https://chipsandcheese.com/2021/09/01/the-weird-and-wacky-world-of-via-the-3rd-player-in-the-modern-x86-market/

[mboeller]

https://chipsandcheese.com/2021/09/01/the-weird-and-wacky-world-of-via-the-3rd-player-in-the-modern-x86-market/

[davidzo]

Nice Article, albeit its all ancient History already.
I can't wait to see Part 2 with the Zhaoxin Architecture analyzed. Does anybody happen to know which CPU they have there, is it a KX-U6780A / LuJiaZui µarch?

I would be most interested in a CentaurCHA / CNS review or µarch analyzing. But i suppose they are pitching it only for their ai coprocessor, not for the CNS cores.

Does anybody know if there is still a Ip transfer between Centaur and Zhaoxin, so we can expect KX7000 to have some of its DNA? or did they go their own route since Isaia II?

[Loeschzwerg]

Nice Article, albeit its all ancient History already.
I can't wait to see Part 2 with the Zhaoxin Architecture analyzed. Does anybody happen to know which CPU they have there, is it a KX-U6780A / LuJiaZui µarch?


Looking at their database entries its a quadcore KX-6000 (LuJiaZui µarch) they have.


Does anybody know if there is still a Ip transfer between Centaur and Zhaoxin, so we can expect KX7000 to have some of its DNA? or did they go their own route since Isaia II?

I assume they still share core IP, looking at the main specs of ZX-F and the identifier (CentaurHauls Family 6 Model 71).

[Tralalak]

Zhaoxin is still Centaur CPUID and microarcitecture.

Last year VIA to sell part of x86 processor IPs to join venture Zhaoxin. The deal is made through VIA's 100%-owned subsidiaries VIABASE (excluding patents) and VIATECH.
source: https://cntechpost.com/2020/10/28/via-sells-technologies-including-x86-to-shanghai-zhaoxin-for-257-million/

[davidzo]

Okay, so KX-7000 Series will most likely be based CNS-µarch.

I wonder what happened to the samples of CHA that Ian was supposed to receive after Dec 2020. Thy must have already been delivered, because market start for the CHA was scheduled for 2021 and he was supposed to get the Samples ahead of that.

From what i read, they seem to position the CHA as an Edge AI Processor very much like Nvidia Jetson Xavier.
In ImageNet it has twice the performance of Xavier, which is impressive, considering it has a similar raw performance in Tops.
But it has also almost twice the memory bandwidth. I wonder if it also lands in around twice the power envelope (40-65W), which would still be good against nvidias tried and tested Volta architecture and Carmel µArch ARM cores.

I don't see any benefit in the x86 ISA when considering the edge AI market. CUDA and CUDA on Arm, Tensorflow, pytorch has been so dominant that this time that Intels AVX512 and Vias "AVX32768" VLIW Approach are the outsiders here that have to do a lot of convincing.

Everybody can just get a Jetson for 59$, put a camera on it, download JetPack SDK and program a basic sceleton detection in less than a week. Thats pretty hard to beat.

I really hope that with Zhaoxin the CNS Core is coming back from AI-servers into general purpose PCs!

[Tralalak]

x86-Exot
Mini-ITX-Mainboard mit chinesischem Prozessor Zhaoxin KX-U6580
source: https://www.heise.de/select/ct/2021/20/2117513415912404118

[mboeller]

second part:

https://chipsandcheese.com/2021/09/22/the-weird-and-wacky-world-of-via-part-2-zhaoxins-not-quite-electric-boogaloo/

[Tralalak]

second part:

https://chipsandcheese.com/2021/09/22/the-weird-and-wacky-world-of-via-part-2-zhaoxins-not-quite-electric-boogaloo/
Please write someone in the discussion. Because I can't. That Tralalak is referring him.
This is not true. "Based on the information on Wikichip, Zhangiang is a minor modification of the Isaiah II core where the biggest changes were the addition of the Chinese hashing algorithms SM3 and SM4 to Padlock."

The true refresh of VIA Isaiah mixroarchitecture CN (CNA, CNB, CNC, CNQ) is VIA Isaiah II - CNR microarchitesture which is equivalent Zhaoxin ZX-C Zhangiang.

[y33H@]

That's actually what they're saying.

[Tralalak]

Please someone write it to him there.
It's based on bad information.
This unnecessarily disqualifies an excellent article.

https://twitter.com/ChipsandCheese9

Thanks

[y33H@]

They say, ZX-C is literally Isiah II, but with SM3/SM4 - and you're saying the same, so everything is fine?

[Tralalak]

VIA CNR - Isaiah II is the biggest change to the original VIA CN - Isaiah (CNA, CNB, CNC, CNQ) microarchitecture.

40nm two die VIA QuadCore CNQ Isaiah vs 28nm VIA QuadCore/Eden X4/Eden X1 CNR Isaiah II or Zhaoxin ZX-C/ZX-C+ Zhangiang
http://abload.de/thumb/l4700ozjqg.jpg vs http://abload.de/thumb/c465090k1y.jpg

VIA CNR - Isaiah II (QuadCore 4650, Eden X4, Eden X1) = equivalent Zhaoxin Zhangiang (ZX-C), ZX-C+ as SM3/SM4 and ZX-C+FC1080 OctaCore

ZX D/KX-5000 Wudaokou is only evolution of VIA CNR - Isaiah II (Zhangiang).

[y33H@]

Yes, so you and Chips/Cheese say the same ...

[Tralalak]

Are you kidding me?
The author of the article does not mention VIA CNR at all.

He thinks that Isaiah II is just die schrink Isaiah (40nm to 28nm) and added only SM3 / SM4 instructions. It's not true!

You made the same mistake!

"Bei den ZX-C alias KX-4000 führte Zhaoxin erstmals Kaixian als Markenname ein, zudem wurde die Isaiah- durch die eigene Zhangjiang-Technik ersetzt. Die CPUID legt nahe, dass hier die Isaiah-2-Technik (CNR) verwendet wird. Offenbar handelt es sich rein um einen Shrink und keine überarbeitete Architektur, denn statt HLMCs 40 nm wurde TSMCs 28 nm herangezogen. Die vier Kerne sind nativ, wie bei Via auch - zuvor gab es auch zwei Chips auf einem Träger. Mit den ZX-C+ koppelte Zhaoxin dann zwei Quadcore-Dies, um so bis zu acht CPU-Kerne zu erhalten. Außerdem wurde die Zhanjiang-Technik um SM3 und SM4 erweitert: Diese Verschlüsselungsverfahren sind erforderlich, damit die Prozessoren in staatlichen Institutionen in China eingesetzt werden dürfen."
source: https://www.golem.de/news/zhaoxin-kx-u6780a-im-test-das-kann-chinas-x86-prozessor-2005-148364.html

[y33H@]

I say, ZX-C's CPU-ID suggests Isiah II ... but let me read that on my Desktop again, Smartphone sucks for checking sources ^^

EDIT

My table is correct, the text on the other hand was (still) buggy - it's edited now.

[Tralalak]

It's not.

[y33H@]

Well, it would be helpful to tell me what's not ...

[Tralalak]

The ZX-C C4000 series has never been called the KX-4000 series.

But that is irrelevant.
The problem is that I can't log in to my twitter account or to their wordpress discussion.

Here is Zhaoxin KaiXian ZX-C (Zhangiang) alias VIA QuadCore/EdenX4 CNR - Isaiah II)

https://i.imgur.com/kOCiF2X.png

https://nimg.ws.126.net/?url=http%3A%2F%2Fdingyue.ws.126.net%2FkJjnjDZoy29HGcluLbgXgStOPdGP2lIvfm610Wmgh J1UH1476409332718compressflag.jpg&thumbnail=650x2147483647&quality=80&type=jpg


vs


evolution die ZX - D WuDaoKou (ZX-E LuJiaZui will be very similar) more L2 cache 4 MB, Improved OoOE algorithm, Pipeline was reduced by 5 stages,
Execution engines were re-balanced, Branch prediction unit was reworked and optimized...

https://en.wikichip.org/w/images/thumb/c/c5/wudaokou_core_%28annotated%29.png/750px-wudaokou_core_%28annotated%29.png

https://i.imgur.com/TUBf1vG.png

P.S. I sent them a message.

[davidzo]

I can feel y33h@ because it is really hard to understand what you mean with all that information.

I suppose the argument is about ZX-C being Isaiah II based, which always was a 28nm native design. so no shrinks here. The article is very misleadingly formulated because it is referring to the predecessors of ZX-C, which are "Isaiah I" based and 40nm at the same time as it is referring to Vias Eden X4/ CNR Isaiah II equivalent and even Vias past "Isaiah I" dual-die variants. The article changes what it is referring to in every sentence and could indeed have different meanings, as it is not clearly recognizable when it is referred to Via and when to the older ZX-A and ZX-B. Also making up a KX4000 Family name for ZX-C and not mentioning ZX-A and ZX-B makes it very complicated as there is no KX3000 and KX2000 also. Indeed you need to know the facts first, to interpret y33h@ the right way, otherwise you would understand it like you said (Isaiah II being a 40nm design first, which is wrong). The translation form german may make it even more clearly misleading.


@y33h@ Vllt. kannst du das hier ändern:


[I]"Bei den ZX-C alias KX-4000 führte Zhaoxin erstmals Kaixian als Markenname ein, zudem wurde die Isaiah- durch die eigene Zhangjiang-Technik ersetzt. An dieser Stelle wäre es präziser auch von "Isaiah II" statt nur Isaiah zu sprechen um keine Missverständnisse hervorzurrufen. Gerade beim Übersetzen ist das Missverständlich, sonst müsstest du ZX-A und ZX-B auch noch erwähnen zusammen mit Isaiah.
Btw, wenn ich richtig verstehe gibt es keine KX-4000 Familie, selbst wenn die modellnamen C/C+ 4000 heißen. Die Familien heißen so wie die internen Codenamen ZX-A bis ZX-C. ZX-A ist Isaiah aus Via produktion, ZX-B ein Isaiah in eigener produktion und ZX-C Isaiah II basiert.
Den KX-Familiennamen hat man erst ab 5000 series, also mit den eigenen Core Architekturen eingeführt.


Die CPUID legt nahe, dass hier die Isaiah-2-Technik (CNR) verwendet wird.
@Tralalak: This part is not actually saying its Isaiah 2 or Isaiah I, it just says that the CPU ID suggests it is based on Isaiah II tech, which is not wrong from what i understand.


Offenbar handelt es sich rein um einen Shrink und keine überarbeitete Architektur, denn statt HLMCs 40 nm wurde TSMCs 28 nm herangezogen. Die vier Kerne sind nativ, wie bei Via auch - zuvor gab es auch zwei Chips auf einem Träger.

@y33h@ Hier kriegst du schon wider Isaiah I und II durcheinander. Isaiah II gab es nie in 40nm, der war schon immer 28nm, insofern ist es falsch hier 40nm zu erwähnen und im gleichen Satz von reinem Shrink zu sprechen. Das hieße ja die µArch wäre noch auf Isaiah I niveau.
Isaiah II gab es auch nie als dualchip variante von Via.
Ich glaube es ist einfach ein Fruchtloses unterfangen in so wenigen Sätzen gleichzeitig den Vergleich zu den äquivalenten Via-µArchs und zu den Vorgängern bei Zhaoxin und den Vorgängern bei Via zu ziehen. Das ist völlig verwirrend.

Übrigens etwas woran leider jeder Golem Artikel irgendwie leidet: Wortkargheit.
Nur ein bisschen mehr Hintergrundinfo. Nur ein bisschen mehr Roadmapbilder, Fotos vom testaufbau (auch gerne Handyfotos etc.) und man wäre nicht mehr weit von Anandtech-niveau entfernt. Aber bisher wirken die Artikel trotz sehr diverser, interessanter Themenwahl leider immer ein bisschen kurz. Manchmal, unzulässig verkürzt, so wie hier.
Interessanterweise macht das die Artikel obwohl sie schön kompakt sind gerade für casual Leser ungeeignet. Denn die können das ohne Hintergrundinformationen nicht verstehen und sind ohne mehr Bilder vollkommen orientierungslos.
Weniger ist manchmal besser und knackiger, aber hier würde ich um etwas mehr bitten. Das bringts imo, denn das Aufwändige ist ja nicht das Texten, sondern der Test an sich, die recherche etc., oder nicht?

[y33H@]

Ich glaube du hast noch die alte Version im Cache, hatte es schon gestern geändert.

Die Kritik der Wortkargheit und noch mehr Hintergrundinfos versuche ich bei kommenden Artikel zu beherzigen, tatsächlich ist der zeitliche Aufwand des Textens aber nicht zu unterschätzen [und generell habe ich gefühlt nie Zeit].

Mich limitiert hier leider auch das Layout unseres CMS oft, da gibt's anderswo schlicht (viel) mehr Möglichkeiten um Elemente wie Bilder/Fotos unterzubringen.

[davidzo]

Ich glaube du hast noch die alte Version im Cache, hatte es schon gestern geändert.


Kann sein, bei mir steht noch das hier:

Bei den ZX-C alias C4000 führte Zhaoxin erstmals das eigene Zhangjiang-Design ein. Die CPUID legt allerdings nahe, dass hier die Isaiah-2-Technik (CNR) verwendet wird. Offenbar handelt es sich um eine verbesserte Architektur plus einen Shrink, statt HLMCs 40 nm wurde TSMCs 28 nm herangezogen. Mit den ZX-C+ koppelte Zhaoxin dann zwei Quadcore-Dies, um so bis zu acht CPU-Kerne zu erhalten. Außerdem wurde die Zhanjiang-Technik um SM3 und SM4 erweitert: Diese Verschlüsselungsverfahren sind erforderlich, damit die Prozessoren in staatlichen Institutionen in China eingesetzt werden dürfen.

- Das klingt immer noch als wenn ZX-C eine Evolution aus ZX-B war. Dem ist aber nicht so. ZX-B und das HMC Abenteuer waren dann zu ende und man ist erneut zu Via gegangen und hat das neueste Design lizensiert und noch dazu genau wie Via selbst im 28nm Verfahren bei TSMC geordert. Da irgendeinen Bezug zu HLMCs 40nm und Isaia I basiertem ZX-B zu ziehen ist einfach zu weit hergeholt.
- Das wäre ja so als wenn ich sagen würde die CPU im Snapdragon 810 basiert auf dem 805, nur in einem neuen prozess und mit Architekturverbesserungen. Die haben architekturell außer der ISA aber nichts gemeinsam, der 805 war qualcomms custom Krait core, und damit eine Sackgasse und der 810 ist ein standard A57.
- ZX-C == Isaiah II, das kann man durchaus so schreiben. Eine "Verbesserte" Arch kann ja alles sein, auch Isaia I mit minimalen Änderungen oder einer neuen befehlssatzergänzung. Die haben da aber gar nichts verbessert, sondern sind einfach zu Via gegangen und haben das neueste design lizensiert. Wenn dann müsste man die Story erzählen wie Via da was an Isaiah II verbessert hat.
- Btw, die erste wirklich eigene µArch die nicht direkt auf einem Via Design basierte, war erst WuDaoKou, nicht schon ZhangJiang. Deshalb wahrscheinlich auch der Namensbeginn mit KX-5000 erst ab ZX- D. ZhangJiang hat sogar den Via Floorplan weitgehend übernommen.
- SM3 und SM4 ist schon im ZX-C drin, nicht erst im ZX-C+ wie im Artikel dargestellt.

[y33H@]

Ich hatte mich an der Roadmap orientiert, die SM3/SM4 erst bei ZX-C+ zeigt.

https://scr3.golem.de/screenshots/2005/Zhaoxin-ZX-U6780A-Review/thumb620/Zhaoxin-ZX-U6780A-Review-08b.png

Hab den Absatz mit dem ZX-C(+) umformuliert, der war nicht präzise bzw missverständlich. Die Tabelle sagt ja klar, dass es Isaia II ist.

[Tralalak]

If you are not tired of my posts, at the end of the summer I had one pleasant interview. (use google translate or another translator)

Čínské x86 procesory na Intel a AMD nemají, pro většinu lidí jsou ale dostačující, říká muž, který je testuje
source: https://www.lupa.cz/clanky/cinske-x86-procesory-na-intel-a-amd-nemaji-pro-vetsinu-lidi-jsou-ale-dostacujici-rika-muz-ktery-je-testuje/

[rasti666]

VIA and Zhaoxin processors have been suitable for gaming for years. I buy all motherboards from them for at least 20 years. I played all the games I wanted. The problem is availability and price. VIA and Zhaoxin are losing to AMD and Intel. VIA and Zhaoxin are in a losing position. The price is too high. I know this - because I bought all their motherboard models. To this day, I use several of them in my company. They are good for the average user. Price and availability is a nightmare. The latest graphics card is the VIA eH1 which I am currently using. I doubt that Zhaoxin will decide to continue graphics cards. I look forward to it. What S3 will I play in Cyberpunk?

[Benutzername]

VIA and Zhaoxin processors have been suitable for gaming for years. I buy all motherboards from them for at least 20 years. I played all the games I wanted. The problem is availability and price. VIA and Zhaoxin are losing to AMD and Intel. VIA and Zhaoxin are in a losing position. The price is too high. I know this - because I bought all their motherboard models. To this day, I use several of them in my company. They are good for the average user. Price and availability is a nightmare. The latest graphics card is the VIA eH1 which I am currently using. I doubt that Zhaoxin will decide to continue graphics cards. I look forward to it. What S3 will I play in Cyberpunk?



That you cannot really get them outside China is a problem for us here yes, but
well, what do the sales numbers in the PRC market look like? Do they have any notable share? Government contracts maybe to keep them afloat to build a domestic chinese CPU maker?

[rasti666]

Of course - it is possible that this is China's "fight" with the rest of the world. Processors are available on the market - so the war is "won". The fact is, production is low.

But, starting with C3, C7-D, VIA Nano, later Quad-Core, and today Zhaoxin KX-U6580 - which I am writing now - there is progress ;). We'll see what will be "tomorrow". I hope there will be more of these processors on the market to our benefit.

[Tralalak]

QNAP TVS-675 review: NAS with special processor
source: https://techacademy.id.nl/blog/qnap-tvs-675-review-nas-met-bijzondere-processor/

[konkretor]

wtf?


https://www.computerbase.de/2021-11/x86-prozessoren-intel-uebernimmt-teile-von-vias-centaur-team/


https://twitter.com/tculpan/status/1456264407101575185


Was hat Intel vor mit dem Design Team von Centaur ?

[Tralalak]

VIA To Offload Parts of x86 Subsidiary Centaur to Intel For $125 Million
source: https://www.anandtech.com/show/17049/via-to-offload-parts-of-x86-subsidiary-centaur-to-intel-for-125-million

[Loeschzwerg]

Was hat Intel vor mit dem Design Team von Centaur ?

Gute Frage. Vielleicht war Intel auf der Suche nach einem alternativen, kleinen Kern Design und da passen ihnen die CNS IP (und oder auch der NCORE) inkl. Teile des Engineering Teams in den Kram.

Klar ist für mich nur eines, VIA geht den gleichen Weg wie andere große Firmen (Siemens, IBM, Fujitsu... usw.), man trennt sich immer mehr vom Hardware Geschäft und geht in Richtung "Service Solution Factory". Finde ich persönlich lahm, aber gut.

Sollte es dann letztendlich das Ende Centaur bzw. deren x86 Entwicklungen und auch Zusammenarbeit mit Zhaoxin (und wiederum deren Produkte) bedeuten, fände ich das extrem schade. Es wurde doch gerade erst wieder etwas spannender ;(

[Benutzername]

Es wird auf jeden Fall ein Konkurrent von intel gekauft und intel hat halt ein weiteres Entwicklungszentrum mit Centaur. Ob überhaupt etwas vom Centaur IP übernommen werden wird, wird man sehen müssen. Müsste das nicht auhc den Kartellbehörden vorgelegt werden in den USA ob der Kauf überhaupt durchgeht?

[Skysnake]

Ich denke eher nicht. Sie kaufen ja nur ein Team und nicht die Firma mit den Rechten

[Tralalak]

Intel will recruit some some of Centaur's employees to join Intel with certain covenants from Company for Intel's recruitment. As consideration, Intel will pay Centaur US$125 Milion.

I assume that Centaur can outsource part of its team for some Intel project.

And in no way can be used to impact VIA's usage of Centaur product lines.

https://www.gvm.com.tw/article/83920

[konkretor]

Ian hat hier noch etwas Stoff

https://www.youtube.com/watch?v=sJm1XmWoVww

[Tralalak]

INTEL WANTS CENTAUR FOR HYBRIDIZATION
That is what Intel is interested in; an HPC coprocessor which can be added to their existing products. Specifically, it is Ponte Vecchio, Intel’s multi-tile hybrid design which will blend a mélange of cores inside a single CPU which could benefit from Centaur’s work. Intel has in the past stated that some of the cores included in Ponte Vecchio will not be designed by Intel, so integrating an HPC coprocessor built on non-Intel architecture would be well within scope.
zdroj: https://pcper.com/2021/11/intel-wants-centaur-for-hybridization/

[robbitop]

Sind für die 125M USD auch die IPs und Patente dabei? Gerade die machen Centaur ja recht wertvoll. Die Designs sind sicher nicht schlecht - aber bis dato verglichen mit dem was Intel/AMD und die ARM Welt so tut eigentlich nur 2. Liga.

[Benutzername]

Sind für die 125M USD auch die IPs und Patente dabei? Gerade die machen Centaur ja recht wertvoll. Die Designs sind sicher nicht schlecht - aber bis dato verglichen mit dem was Intel/AMD und die ARM Welt so tut eigentlich nur 2. Liga.


To me this looks like intel hiring more engineers without Centaur's patents and stuff. This way they get more manpower and circumvent any monopoly rules, becuase Centaur still exists. Doesn't matter if it is an empty husk.


Makes me wonder how dependent ZhaoXhin is on Centaur nowadays? Or can they stand on their own making complete CPUs?