3 ms·

pokes OpenCL's corpse with a stick C'mon, do something...

ein0p 3y ago
Ironically, Transformers are relatively simple architectures - all you really need is a high performance matmul. So OpenCL could "do something" at this point, if it were alive.
imtringued 3y ago
https://github.com/ROCm/ROCm/issues/2754 https://github.com/ROCm/ROCm/issues/2754 Wow and I thought that the latest generation of GPUs was better.
- anakaine 3y ago
  From that exchange, clearly not. They do not have their stuff together. This is one area where AMD really should be employing quite a full team to ensure that major projects, including drivers, are able to support AMD cards without any friction. Whilst it's still a PITA to work.with AMD hardware in most AI frameworks Nvidia will continue to dominate.
anakaine 3y ago
Its very unfortunate. One of our vendors has spent a hell of an effort making spatial analysis tools work via OpenCL instead of being just processor bound, and its made those industry standard libraries 1-2 orders of magnitude faster. That is something the spatial industry desperately needs in order to improve iteratively. With the state of OpenCL it's frustrating. So much to be had, but so little improvement and support.
- talldayo 3y ago
  In 50 years, I feel like GPGPU compute will be told as a Greek tragedy. Nvidia, pariah of Apple, would create a parallel compute ecosystem that was so powerful that even the combined effort of the industry couldn't topple it. At every corner where they could have thwarted them, Nvidia's competitors refused to up the ante or chided CUDA's efforts as silly and unnecessary. One crypto/AI craze later, everyone and their mother is berating their OEM for ignoring such basic functionality and refusing to collaborate on a competing alternative. Frustrating is a good way to put it, but there's a certain causal satisfaction I feel from watching it all unfold. Of course everyone loses to CUDA when they refuse to sponsor an Open Source alternative. It's fascinating to me that hardware manufacturers would rather let CUDA dominate than establish a basic working relationship.

Ironically, Transformers are relatively simple architectures - all you really need is a high performance matmul. So OpenCL could "do something" at this point, if it were alive.

imtringued 3y ago

https://github.com/ROCm/ROCm/issues/2754 https://github.com/ROCm/ROCm/issues/2754 Wow and I thought that the latest generation of GPUs was better.

anakaine 3y ago
From that exchange, clearly not. They do not have their stuff together. This is one area where AMD really should be employing quite a full team to ensure that major projects, including drivers, are able to support AMD cards without any friction. Whilst it's still a PITA to work.with AMD hardware in most AI frameworks Nvidia will continue to dominate.

anakaine 3y ago

Its very unfortunate. One of our vendors has spent a hell of an effort making spatial analysis tools work via OpenCL instead of being just processor bound, and its made those industry standard libraries 1-2 orders of magnitude faster. That is something the spatial industry desperately needs in order to improve iteratively. With the state of OpenCL it's frustrating. So much to be had, but so little improvement and support.

talldayo 3y ago
In 50 years, I feel like GPGPU compute will be told as a Greek tragedy. Nvidia, pariah of Apple, would create a parallel compute ecosystem that was so powerful that even the combined effort of the industry couldn't topple it. At every corner where they could have thwarted them, Nvidia's competitors refused to up the ante or chided CUDA's efforts as silly and unnecessary. One crypto/AI craze later, everyone and their mother is berating their OEM for ignoring such basic functionality and refusing to collaborate on a competing alternative. Frustrating is a good way to put it, but there's a certain causal satisfaction I feel from watching it all unfold. Of course everyone loses to CUDA when they refuse to sponsor an Open Source alternative. It's fascinating to me that hardware manufacturers would rather let CUDA dominate than establish a basic working relationship.