LibreCUDA – Launch CUDA code on Nvidia GPUs without the proprietary runtime
1–10 of 162 posts
Re: LibreCUDA – Launch CUDA code on Nvidia GPUs without the proprietary runtime
#2Very nice! That's essentially all I want from a cuda runtime. It should be possible to run llvm libc unit tests against this, which might then justify a corresponding amd library that does the same direct to syscall approach.
Re: LibreCUDA – Launch CUDA code on Nvidia GPUs without the proprietary runtime
#3For a non cuda n00b, what problem does this solve?
Re: LibreCUDA – Launch CUDA code on Nvidia GPUs without the proprietary runtime
#4For a non cuda n00b, what problem does this solve?
Like anything open source it allows you to know and see exactly what your machine is doing. I don’t want to speculate too much but I remember there being discussions around whether or not nvidia could embed licensing checks and such at the firmware level.
Re: LibreCUDA – Launch CUDA code on Nvidia GPUs without the proprietary runtime
#5Do you still need to be running the proprietary nvidia graphics driver, or is that completely unrelated?
Re: LibreCUDA – Launch CUDA code on Nvidia GPUs without the proprietary runtime
#6For a non cuda n00b, what problem does this solve?
Two obvious problems that come to mind are
1. Replacing the extremely bloated official packages with lightweight distribution that provides only the common functionality.
2. Paving the way for GPU support on *BSD.
Re: LibreCUDA – Launch CUDA code on Nvidia GPUs without the proprietary runtime
#7What's a CUDA elf file?
Is it binary SASS code, so one would still need a open source ptxas alternative?
Re: LibreCUDA – Launch CUDA code on Nvidia GPUs without the proprietary runtime
#8Since ZLUDA was taken down (by request from AMD of all parties), it would be better to have some ZLUDA replacement as a general purpose way of breaking CUDA lock-in. I.e. something not tied to Nvidia hardware.
Re: LibreCUDA – Launch CUDA code on Nvidia GPUs without the proprietary runtime
#9Since ZLUDA was taken down (by request from AMD of all parties), it would be better to have some ZLUDA replacement as a general purpose way of breaking CUDA lock-in. I.e. something not tied to Nvidia hardware.
That's a problem on a different level of the CUDA stack.
Having a compiler that takes a special C++ or python dialect and compiles it to GPU suitable llvm-ir and then to a GPU binary is one thing (and there's progress on that side: triton, numba, soonish mojo), being able to launch that binary without going through the nvidia driver is another problem.
Re: LibreCUDA – Launch CUDA code on Nvidia GPUs without the proprietary runtime
#10Do you still need to be running the proprietary nvidia graphics driver, or is that completely unrelated?
Presumably yes, if it functions through an ioctl interface.