3 ms·
I remain a fanatically committed AMD buyer for their linux support, but wow the cost of it in the machine learning space has been painful. It really hurts when
by roenxi 4y ago
I remain a fanatically committed AMD buyer for their linux support, but wow the cost of it in the machine learning space has been painful. It really hurts when all the tutorials are for Nvidia and nothing seems to work.
I suspect the ROCm situation has not improved since last time I checked, based on their documentation. The installation guide [0] is a bit of a mess. It looks like they are targeting ROCm at existing ROCm users rather than finding ways to convince people to become ROCm users; nothing there is useful introductory material. The opening is basically marketing babble too. Or possibly someone with a history in the GNU projects lineage of terrible recursive names.
It just doesn't look like they've got a leadership figure trying to bring people in to their ecosystem right now. I dunno, maybe the API isn't intended for interested developers.
[0] https://docs.amd.com/bundle/ROCm-Installation-Guide-v5.4/page/Introduction_to_ROCm_Installation_Guide_for_Linux.html https://docs.amd.com/bundle/ROCm-Installation-Guide-v5.4/pag... - opens with "ROCm is a brand name for the ROCm open software platform (for software) or the ROCm open platform ecosystem (includes hardware like FPGAs or other CPU architectures)".
- rerx 4y agoThe tooling is there to convert existing CUDA code into a form that runs on AMD GPUs as well: https://docs.amd.com/en-US/bundle/HIPify-Reference-Guide-v5.1/page/HIPify.html https://docs.amd.com/en-US/bundle/HIPify-Reference-Guide-v5.... They have drop-in replacements for most Nvidia-specific libraries like cuDNN, too. So, while the documentation still doesn't make it trivial to get started, the software side definitely seems to have improved in recent years.
- Cloudef 4y agoIt's unfortunate. CUDA is a mess, but because it's the leading tech on the field, AMD has no choice but to come with an (equally messy) compatible solution. Vulkan was thought of as a something that could replace CUDA and (dead) OpenCL, but it's never going to take off due to the ML field being heavily CUDA, or well more of pytorch / python :)
- eyegor 4y agoI haven't tried it myself, but pytorch has supported amd gpus via rocm since 2021. https://www.amd.com/en/technologies/infinity-hub/pytorch https://www.amd.com/en/technologies/infinity-hub/pytorch
- berkut 4y agoI'd personally argue ROCm is much more of a mess than CUDA: in many situations, CUDA code can be compiled once, and then run on multiple generations of cards (including future ones once drivers exist) and you have the option of shipping PTX code which can (normally) do the same. ROCm builds are effectively hardware-specific (and officially Linux-only).