Windows 11 Better Than Linux Right Now For Intel Alder Lake Performance

[Scellow]

Very fishy results.. i wouldn't say there is a sabotage somewhere, but i'm close

As other pointed out, take a look at this interesting article:

https://www.anandtech.com/show/16881/a-deep-dive-into-intels-alder-lake-microarchitectures/5

Could it be related to the AVX-512 instructions, is there some sort of special treatment on windows? is that even possible?

The gap on the selenium benchmark is kinda suspicious

[Termy]

I think it was Intel that did the work .

It's hard for me to believe MS would give source access to outside parties tbh...

[smitty3268]

It's hard for me to believe MS would give source access to outside parties tbh...

They've done so for quite some time. I think governments were the first to demand access to the source code, for obvious security reasons.

Now there's a whole program around it, and it doesn't sound like it's that difficult to qualify.

https://www.microsoft.com/en-us/sharedsource/

Edit: That doesn't mean I think Intel did the work, though. I think it was still Microsoft, though they probably worked with Intel developers together.

[krzyzowiec]

That's why my new PC, that will arrive next month if shipping isn't delayed somehow, has an AMD Ryzen 9 5900X.

My first AMD in, like, forever. (the last time I had one was many years ago when the proprietary Radeon drivers were always slow and broken on Linux)

You will love it. I’ve had it since release and I’ve been blown away by how quick and efficient it is. It just feels like overkill to me now. I actually would rather wait like five or ten years for something special on my next upgrade, like a new architecture or something, because I just don’t see myself needing more performance for a long while.

[avem]

So the alder lake drivers for linux are gimped? good to know.

There's no such thing as "drivers" for the CPU.

At most we're talking about the code to deal with the instruction set and the scheduler (which is architecture agnostic) to choose proper cores for running tasks - the latter is actually what the issue is about.

[avem]

So I guess Intel's work on Linux is just for show and keeping the community alive, but nothing more.
There's always *one* feature that Intel does on Windows first, for whatever reason.

Even worse, they got cooler names on Windows (Quick Sync, Turbo Boost, Thread Director, etc.), while what do we get on Linux? VA-API, intel_pstate, i915...

Turbo Boost doesn't need OS support to work but yes their Windows drivers are leaps and bounds better than Linux' just like AMD's and NVIDIA's.

[microcode]

Heterogenous scheduling is a problem like modern fighter jet control software: the thing is inherently more wacky and unstable, but can perform amazing feats if tuned correctly.

[lucrus]

Insecure? Cant remember the last time i had a virus on my pc since 28 years. No one knows more about security than Microsoft, given their history and constant threat to Windows. Windows is built like an armored tank.

Sure. I can't remember the last time I had AIDS since I was born, so I can deduce AIDS does not exist anymore for anyone, right?

Besides, you can't tell anything about what MS Windows team actually knows about security, because their code is secret, their coding practices are secret and the only public information we have is the fact their product (Windows) is constantly a target for viruses and the like. Maybe there is a solid reason if virus writers mainly target Windows...

[torsionbar28]

Intel has shown that the linux desktop is completely irrelevant.

Nailed it. Linux is optimized for commercially viable platforms. I.e. server and embedded devices. We will not see Alder Lake optimization until early 2022 at the soonest when Sapphire Rapids Xeon launches. Desktop users only get to reap the benefit if/when the desktop chip shares uarch with the server chip. If you're that committed to Linux on desktop use Ryzen instead (shares uarch with EPYC) or use the Xeon E series. Personally I run a Xeon E-2276G running Fedora for serious work, and a Ryzen 3600 for Linux gaming. Alder Lake is stuck in an odd spot at the moment.

[agd5f]

Honestly, comments like this make me wonder why there apparently isn't a proper communication channel inside AMD between different divisions?

You say that the decision-making logic should be left up to the hardware to decide, while schedutil proponents argue that the hardware can't possibly have a clue about OS run-time queues of all the different threads interacting with each other.

I'm not sure what you are asking. Linux and windows take different approaches to CPU clock management in my understanding. Both OSes still give hints to the the hardware. CPPC (the underlying interface to control the CPU clocks) was designed for windows and seems to work well there. Linux seems to be more hands on while windows less so. I suspect windows gives target hints at a pretty coarse level and then lets the hardware go (here's my target performance, with that in mind, get it done as fast as possible) while Linux seems to constantly be setting new targets for performance (more work coming online, lets try a slightly faster target, now there's less work, let's try and slow things down). With the old pstate APCI interface, there were only 3 states, so even if the OS was constantly setting new targets, you'd just end up snapping to the nearest state. CPPC gives you a continuum of performance states, so every time you set a new hint, you could potentially end up walking through a long continuum of frequencies. I'm certainly not an expert in this area, just seems that way from what I've seen.

As stated by others on the thread, the desktop is not necessarily the main use case for Linux at this point. I suspect the current schedulers work well for embedded and server platforms. In those cases you might favor something more deterministic which it seems like the Linux schedulers strive for.