Alternative(s) to run CUDA on non-Nvidia hardware
1–10 of 82 posts
Re: Alternative(s) to run CUDA on non-Nvidia hardware
#2Most of these "alternatives" focus on CUDA C++, and overlook what actually makes CUDA interesting.
Already in 2020,
https://developer.nvidia.com/blog/cuda-refresher-the-gpu-com...
Re: Alternative(s) to run CUDA on non-Nvidia hardware
#3Most of these "alternatives" focus on CUDA C++, and overlook what actually makes CUDA interesting. Already in 2020, https://developer.nvidia.com/blog/cuda-refresher-the-gpu-com...
We're actually targeting all of it, and not just CUDA C++.
Re: Alternative(s) to run CUDA on non-Nvidia hardware
#4Most of these "alternatives" focus on CUDA C++, and overlook what actually makes CUDA interesting. Already in 2020, https://developer.nvidia.com/blog/cuda-refresher-the-gpu-com...
We're actually targeting all of it, and not just CUDA C++.
Including stuff like Fortran, Haskell, Java, .NET via PTX, Python JIT, IDE tooling integration with major IDEs, graphical GPU debugging and profiling, libraries and co?
Then I guess all the best.
Re: Alternative(s) to run CUDA on non-Nvidia hardware
#5Most of these "alternatives" focus on CUDA C++, and overlook what actually makes CUDA interesting. Already in 2020, https://developer.nvidia.com/blog/cuda-refresher-the-gpu-com...
We're actually targeting all of it, and not just CUDA C++.
Ambitious but neat, good luck if nothing else :)
If you were to guess, when do you think your Nsight Compute alternative might be ready with your own toolchain?
Re: Alternative(s) to run CUDA on non-Nvidia hardware
#6There is also ZLUDA, which is open source and works on pre-compiled binaries.
Re: Alternative(s) to run CUDA on non-Nvidia hardware
#7Its easier to just get rid of your legacy code entirely and use Vulkan for compute, or have your compiler emit SPIR-V directly.
No reason to tie yourself to Nvidia's moat.
Re: Alternative(s) to run CUDA on non-Nvidia hardware
#8I have been trying for cuda -> metal, to run it on mac, https://github.com/lulzx/cuda-metal
Re: Alternative(s) to run CUDA on non-Nvidia hardware
#9Earlier quoted context omitted.
We're actually targeting all of it, and not just CUDA C++.
Including stuff like Fortran, Haskell, Java, .NET via PTX, Python JIT, IDE tooling integration with major IDEs, graphical GPU debugging and profiling, libraries and co? Then I guess all the best.
This post has some serious peanut-gallery vibes.
Re: Alternative(s) to run CUDA on non-Nvidia hardware
#10Its easier to just get rid of your legacy code entirely and use Vulkan for compute, or have your compiler emit SPIR-V directly. No reason to tie yourself to Nvidia's moat.
[dead]