Quoting: GrabbyYou're probably correct on all those points. But gradyvuckovic's remark was more about Nvidia realizing that a lot of good things can come out an open-source ecosystem.

Nvidia doesn't care. They are like Oracle in this sense. They deal with open source very reluctantly, because they totally don't get the point of it.

View this comment - View article - View full comments

Quoting: MadeanaccounttocommentBut Nvidia's proprietary driver very likely doesn't incorporate LLVM and has probably had a gaming specialized shader compiler from the start.

Nvidia's compiler is using llvm if I remember correctly. At least in some cases.

Their llvm backend: https://github.com/llvm-mirror/llvm/tree/master/lib/Target/NVPTX

Not sure if their new blob Vulkan compiler is using it though.

UPDATE:

Related: https://www.phoronix.com/scan.php?page=article&item=nvidia-396-nvvm

See also: https://docs.nvidia.com/cuda/nvvm-ir-spec/index.html

QuoteNVVM IR is a compiler IR (internal representation) based on the LLVM IR. The NVVM IR is designed to represent GPU compute kernels (for example, CUDA kernels). High-level language front-ends, like the CUDA C compiler front-end, can generate NVVM IR. The NVVM compiler (which is based on LLVM) generates PTX code from NVVM IR.

So, likely they do use llvm to translate PTX into machine code.

View this comment - View article - View full comments

Quoting: gradyvuckovicI'm loving their commitment more than anything. Most companies would have given up by now.

It probably helps, that Valve is a private company, so they can work on long term investments. Public ones are pressured by external investors to rip something off the market quick, and not to care about long term progress.

View this comment - View article - View full comments

I think their plan is to get to the point where it's merged into upstream Mesa, so everyone will get it.

View this comment - View article - View full comments

Quoting: GuestA whole list of problems like 1.

Not sure what your point is. That one issue (not upstreamed driver) is causing everything else. I already explained it above. If you and Nvidia don't care about the progress of the Linux desktop (i.e. Wayland), it doesn't mean Nvidia supports Linux well. It means exactly that - they don't care. The result of not caring is usually garbage.

Quoting: ShmerlMesa is still lagging behind.

How exactly? Mesa is ahead of Nvidia blob in actual compliance to both OpenGL and Vulkan and it caught up in performance. So what is behind?

View this comment - View article - View full comments

Quoting: jensMy best guess would still be a Google Stadia alternative.

Since they also focus on VR heavily, it could be something else too. VR doesn't really fit well with Stadia-like use case.

View this comment - View article - View full comments

Quoting: YoRHa-2BYou cannot multi-thread compilers without sacrificing performance of the generated code. It's just not possible.

That really very much depends on the use case, and how compiler is designed and what it actually does. I'm not sure about shaders use case, but if some stages of the compilation are independent, they can be parallelized in general. That's quite commonly used in compilers to make compilation itself faster.

Example: https://internals.rust-lang.org/t/parallelizing-rustc-using-rayon/6606

View this comment - View article - View full comments

Quoting: YoRHa-2BI've had access to this for several months now and it really is promising. Most of my games do see small improvements, enough to match native D3D11 performance in some cases.

And then there's Nier: Automata, which sees a massive improvement

Is the actual compiler multithreaded? I actually felt stronger stuttering when shaders were recompiled for the first time. Resulting fps is higher, but to reduce stuttering, shouldn't it spread CPU load better?

View this comment - View article - View full comments

Quoting: KristianDo any AMD cards have any kind of ray tracing acceleration?

No dedicated hardware in the current models, AMD recommend using general GPU features for that. But they plan to add something in their future cards (probably to due to pressure from Nvidia).

View this comment - View article - View full comments

Quoting: MohandevirI'm willing to give AMD another try... In another brand. :)

I have very good experience with Sapphire.

View this comment - View article - View full comments