Intel Arc Graphics A770: Windows 11 vs. Linux Benchmarks

[brucethemoose]

One thing I really want to see is "real world" AI benchmarks like Stable Diffusion and ESRGAN on the A770.

Its cheap for a 16GB card, and theoretically powerful, but whether its really as fast as a 12GB RTX 3060 is anyone's guess.

https://github.com/rahulunair/stable_diffusion_arc

... and I cant even find an OpenVINO upscaling example, you would probably have to modify existing example code.

[NotMine999]

Ok.

So the Intel A770 graphics card needs more work on Linux...but if you have been reading Phoronix for the past 6 months or so you already knew that.

Thus, nothing to whine about here folks. Move along.

Thanks for the test info Michael.

[TemplarGR]

Sorry to say but this is not a proper comparison. On the Linux side, wrappers are used for D3D11/12 to Vulkan... Not to mention any WINE overhead... It is useful as a tool to know if anyone can play those games on an Intel gpu running Linux, but is not a real comparison between the 2 drivers. You should have used games that have a native Linux version, or at the very least OpenGL/Vulkan only. I am not too familiar with Strange Brigade, but IIRC it has a Vulkan API mode, was this used in this test?

[dimko]

Well, that was a pile of steaming bull****. Something tells me they are so busy with Windows drivers, that Linux performance/bug fixes are not gonna see light of day any time soon.

[tildearrow]

And this is why I want proprietary Intel drivers for Linux!!!

AMD offers them, and they saved the month during the early Vega days.

[Grinness]

Michael
When can we see benchmarks without the E/P core b******t?
Or at least see how performance compare on a (say) AMD equivalent (anything from 3xxxx to 7xxxx CPU)

Why compute workload are equivalent Linux/windows and gaming is linux 60% of Windows (exactly same story -- or very similar -- when you compare AMD GPU and Nvidia GPU Linux vs Windows)?

EDIT:
Today I checked HW Unboxed video on RTX 4070 ti:

https://www.youtube.com/watch?v=1mE5aveN4Bo

They use a AMD 5800 3d (see minute 2:48) -- Windows OS
They report 81 FPS average for a rx 6800, Cyberpunk 1440p High Quality (see minute 9:42)

On my end, Archlinux r9 5900x and rx6800: 85 FPS average (1440p, High Quality no FSR -- with FSR I get over 99 FPS average)

Again, something odd in your setup

(edit: fixed typos)

[lyamc]

Michael
Why compute workload are equivalent Linux/windows and gaming is linux 60% of Windows (exactly same story -- or very similar -- when you compare AMD GPU and Nvidia GPU Linux vs Windows)?

I have a few guesses, but if it's a case of the Intel Engineers spent most of their time working on the Windows drivers, then we're looking at the 80/20 rule or something similar. Anyways, I'm still not complaining, I got this card for cheap, about $100 less than an equivalent Green or Red.

[sarmad]

And this is why I want proprietary Intel drivers for Linux!!!

AMD offers them, and they saved the month during the early Vega days.

What does proprietary have to do with performance? AMD provided proprietary because they already had it before switching to the open source one, not because optimizing a proprietary driver is magically easier than optimizing an open source driver.

[tildearrow]

What does proprietary have to do with performance? AMD provided proprietary because they already had it before switching to the open source one, not because optimizing a proprietary driver is magically easier than optimizing an open source driver.

Have you seen benchmarks from the early days? AMDGPU-PRO outperformed Mesa for a couple months until Mesa caught up and began outperforming the proprietary driver.

[Grinness]

I have a few guesses, but if it's a case of the Intel Engineers spent most of their time working on the Windows drivers, then we're looking at the 80/20 rule or something similar. Anyways, I'm still not complaining, I got this card for cheap, about $100 less than an equivalent Green or Red.

The issue is another one.

The numbers posted by Phoronix (for AMD, NVidia -- and possibly Intel -- linux vs windows) are not credible.
As I wrote above , my linux system goes neck on neck (if not better) with a similar windows system (AMD linux all open-source driver stack vs AMD closed source windows)
Bero Tech you tube channel (https://www.youtube.com/@beronori/videos) shows the same with his system (AMD 5800x 3d and rx6700xt)

If the problem with Phoronix is the intel E/P scheduler, Ubuntu, the lack of control in the automated scripts or something else I do not know.
What I know is that (with all respect to Michael work) Phoronix credibility is seriously under question