News for hackers. Stuff that matters.
Re:Tensorflow 0.6.0 Release by fizixer on Wednesday December 09, 2015 @07:33PM (#10708018)
Need OpenCL support. Please don't contribute to maintaining CUDA monopoly.
Re:Tensorflow 0.6.0 Release by nl on Thursday December 10, 2015 @12:13AM (#10709014)
If you compare the effort that AMD spends on OpenCL compared to what NVidia spends on CUDA then you'll see why everyone just used NVidia.

I'm not a big fan of vendor "standards", but I have very limited sympathy for OpenCL here.

I think the best hope for portability is at the higher level programming API layer. For example TensorFlow is careful to make switching between CPU and GPU painless.

Re:Tensorflow 0.6.0 Release by oneofthose on Thursday December 10, 2015 @12:46PM (#10712017)
It would not have to be like that if Nvidia opened up the source-code for cuFFT/cuDNN/cuBLAS. My guess is they are not doing that because it is fairly trivial to port code from CUDA to OpenCL. It can even be automated.
Re:Tensorflow 0.6.0 Release by argonaut on Wednesday December 09, 2015 @07:50PM (#10708089)
Unfortunately, that ship sailed years ago. Also, NVIDIA is heavily investing in deep learning (cudnn), so this won't happen unless someone in OpenCL builds something equally/more performant.
Re:Tensorflow 0.6.0 Release by frozenport on Wednesday December 09, 2015 @08:13PM (#10708189)
Somebody like Google?
Re:Tensorflow 0.6.0 Release by argonaut on Wednesday December 09, 2015 @08:49PM (#10708312)
That could work, but Google doesn't seem to have a stake in OpenCL (so it's not a priority). Nvidia has a stake in CUDA, so I wouldn't be surprised if it's a top business priority there.
Re:Tensorflow 0.6.0 Release by dharma1 on Thursday December 10, 2015 @05:10AM (#10709847)
This supports OpenCL (Samsung needed it for TV's).

https://velesnet.ml/

However the benchmarks for OpenCL look about 5x slower than CUDA

Re:Tensorflow 0.6.0 Release by sagargv on Thursday December 10, 2015 @02:54AM (#10709438)