CUDA Dominance vs AMD
The cluster focuses on NVIDIA's CUDA ecosystem dominance in GPU computing, particularly for AI/ML workloads, and AMD's challenges in developing competitive alternatives like ROCm, HIP, OpenCL, and Vulkan.
Activity Over Time
Top Contributors
Keywords
Sample Comments
How is AMD addressing CUDA dominance?
It's not Nvidia's fault that the competition (AMD) does not provide the right software. There is an open alternative to CUDA called OpenCL.
AMD GPUs are better than Nvidia for that - they had async compute way longer. I thought the problem was CUDA lock-in and lack of a nice programming model using something modern like Rust may be.
CUDA is a vendor lock-in scheme. Use OpenCL or Vulkan instead (yes, Vulkan includes support for compute, not just graphics!). AMD supports both, in addition to tools like HIP to help you port legacy CUDA code.
It'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 :)
The market desperately needs an alternative to CUDA, but I just don't see AMD doing it
This is what many people outside the AI world don’t seem to understand. Nvidia has a stranglehold in the form of CUDA and Cudnn. There isn’t any open source equivalent to Cudnn. AMD is trying to push OpenCl in this direction but it will be a long time before DL libraries start migrating to OpenCl. Like tomorrow by miracle if al alternative GPU which is as good as the 1080ti popped up, it would be useless in the AI market.
CUDA is huge and nvidia spent a ton in a lot of "dead end" use cases optimizing it. There have been experiments with CUDA translation layers with decent performance[1]. There are two things that most projects hit:1. The CUDA API is huge; I'm sure Intel/AMD will focus on what they need to implement pytorch and ignore every other use case ensuring that CUDA always has the leg up in any new frontier2. Nvidia actually cares about developer experience. The most prominent exa
I think this is primarily due to the immense effort that NVIDIA has put into CUDA. It works very well and it is extremely fast. The alternatives for AMD are OpenCL and ROCm which have seriously lagged behind CUDA in every respect.EDIT: lots of theories and discussion here: https://www.reddit.com/r/MachineLearni
Ping me when the software stack for the AMD hardware is as good as CUDA.