Steam For Linux In July Shows A 1.23% Marketshare, AMD CPUs Now More Common Than Intel On Linux

[Espionage724]

Soon Linux will finally overtake Apple's macOS!

But frankly, this is expected, since objective benchmarks such as GravityMark show what a joke even Apple's very expensive M1 Max is, even when making use of its tailor-made Metal API & running natively on ARM:

Mac Studio (M1 Max, 2022)
Score: 15,114
https://gravitymark.tellusim.com/rep...c7e577404ca30d

AMD Radeon RX 5600 XT (with RADV on Linux)
Score: 16,467
https://gravitymark.tellusim.com/rep...ceffc49649996b

NVIDIA GeForce RTX 2070 (Mobile)
Score: 17,963
https://gravitymark.tellusim.com/rep...a4c036734bb5f9

Meanwhile Apple fanatics come up with headlines such as this one:

"Apple's M1 Max GPU is as powerful as an Nvidia RTX 2080 desktop GPU and the Sony PS5 gaming console"

https://www.reddit.com/r/apple/comments/qb5ozq/apples_m1_max_gpu_is_as_powerful_as_an_nvidia_rtx/

Am I missing something? A benchmark score in the tens of thousands, with a difference of only about 2000 between an in-house recent GPU vs decent discrete GPUs shows the M1 Max being appealing. I'd easily choose the M1 Max Mac over the other two options since it holds its ground comparably, while also having the advantages of Apple and macOS.

[HEL88]

Windows growth also pretty good + 0.66%, to above 97% :O. Nice.

Windows as total dominator came back .

[bple2137]

I really wish Valve would hurry up and publish an official 3.0 ISO.

I don't think they should. It will be just as good (or bad) as any other desktop distro. For now, they heavily rely on Gamescope for their gaming mode, which is Wayland compositor and doesn't work on NVIDIA GPUs (or barely works). I don't think that people will rush replacing their Windows OSes with Linux just because of Valve's branding, and if so, they'll be disappointed. Also, Valve makes SteamOS primarily as an OS for OEMs to put on their hardware, which makes much more sense than releasing another general purpose Linux distro for enthusiasts.

[Linuxxx]

Wow, 90FPS vs 107FPS. That's such a massive difference... Remind me what the TDP of that M1 Max chip is, vs those two graphics cards again?

I was thinking more about price-to-performance ratio, which is why I wrote 'very expensive'.

In fact, I just had a look:

The cheapest Mac Studio with the M1 Max costs "only" 2.299,00 € in the part of Europe where I live.

Just for comparison, I paid exactly 1111 € for my OEM-PC with an Intel i7-11700F + nVidia RTX 3060 Ti, so actually less than half of the cheapest Mac Studio.

And here's my GravityMark score with a slightly lower resolution, but otherwise exactly the same settings:
(But you got to give me bonus points for running the Windows version through WINE-Staging, plus translating the Direct3D11 calls to Vulkan via DXVK! )



System was Ubuntu Budgie 22.04 LTS running in OpenBox, because the Mutter-based compositor exhibited problems with unlocked framerates.

So, I think it's fair to say that my machine provides the better value, wouldn't you agree?
[But to be honest, I don't care about your opinion -- at all... ]

[hamishmb]

You're welcome. It's rare and quite unusual to see people thanking me here. Most of the time instead it's "You're a bloody troll, you hate Linux, you love NVIDIA/Intel" none of which is even remotely true.

Yeah, I think we would all do well to generally be more polite and receptive here, I'm sorry that that happens to you.

Besides, there's not anything inherently wrong with liking NVIDIA, Intel, or Windows anyway.

[Ladis]

Well, Apple is steadfast in refusing to assign a TDP value on their silicon, so it's N/A vs 185W RX 5600 XT and 115W RTX 2070 Mobile.

Because TDP just means what the chip maker officially says to the device maker to cool down. Nowadays you, as the customer, choose by the specific device's ability to cool down in long term tasks, not by the chip inside. Since Apple is both the chip maker and device maker, there's no point to tell the TDP between those. For you, as a customer, you care only about the power consuption (probably of the whole device for your specific task(s)) and long term performance (what performance it manages to cool down for long term).

[birdie]

Those numbers are incredible, thanks for sharing.

You're welcome. It's rare and quite unusual to see people thanking me here. Most of the time instead it's "You're a bloody troll, you hate Linux, you love NVIDIA/Intel" none of which is even remotely true.

[hamishmb]

Apple M1 Max measured TDP in GPU workloads is around 70W. The whole package can eat up to 120W.

It's wildly more power efficient than RTX 2070 mobile and RX 5600 XT.

Source: https://www.anandtech.com/show/17024...mance-review/3

Those numbers are incredible, thanks for sharing.

[birdie]

Apple M1 Max measured TDP in GPU workloads is around 70W. The whole package can eat up to 120W.

It's wildly more power efficient than RTX 2070 mobile and RX 5600 XT.

Source: https://www.anandtech.com/show/17024...mance-review/3

[hamishmb]

Soon Linux will finally overtake Apple's macOS!

But frankly, this is expected, since objective benchmarks such as GravityMark show what a joke even Apple's very expensive M1 Max is, even when making use of its tailor-made Metal API & running natively on ARM:

Mac Studio (M1 Max, 2022)
Score: 15,114
https://gravitymark.tellusim.com/rep...c7e577404ca30d

AMD Radeon RX 5600 XT (with RADV on Linux)
Score: 16,467
https://gravitymark.tellusim.com/rep...ceffc49649996b

NVIDIA GeForce RTX 2070 (Mobile)
Score: 17,963
https://gravitymark.tellusim.com/rep...a4c036734bb5f9

Meanwhile Apple fanatics come up with headlines such as this one:

"Apple's M1 Max GPU is as powerful as an Nvidia RTX 2080 desktop GPU and the Sony PS5 gaming console"

https://www.reddit.com/r/apple/comments/qb5ozq/apples_m1_max_gpu_is_as_powerful_as_an_nvidia_rtx/

I think Apple's graph at launch is probably responsible for that - it was very misleading.

That said, that a low power SoC comes anywhere near an RTX 2070 mobile and RX 5600 XT is impressive to me.