News for hackers. Stuff that matters.

A comment on Intel Announces Knights Mill: A Xeon Phi for Deep Learning. Parent comment

Re:Intel Announces Knights Mill: A Xeon Phi for Deep Learning by argonaut on Wednesday August 17, 2016 @08:02PM (#12309482)
But being drop-in compatible means supporting CUDA or the CPU interface.
[ Parent | on HN ]
Re:Intel Announces Knights Mill: A Xeon Phi for Deep Learning by cs702 on Wednesday August 17, 2016 @08:06PM (#12309503)
They could also contribute code to the most popular frameworks, instead of releasing forks like "Intel Caffe."[1]

[1] https://github.com/intelcaffe/caffe

[ Parent | on HN ]
Re:Intel Announces Knights Mill: A Xeon Phi for Deep Learning by argonaut on Thursday August 18, 2016 @03:04AM (#12310844)
Yes, but unless the framework authors maintain the contribution (possibly unlikely depending on the situation) they will have to keep maintaining those contributions.
[ Parent | on HN ]
Re:Intel Announces Knights Mill: A Xeon Phi for Deep Learning by joe_the_user on Thursday August 18, 2016 @12:06AM (#12310379)
Is there a problem with a system that's CUDA compatible?

CUDA seems to me like the only simple SIMD-type computing system that's fairly straightforward to program and understand at this point.

Drop-in CUDA compatibility seems like a good thing.

[ Parent | on HN ]
Re:Intel Announces Knights Mill: A Xeon Phi for Deep Learning by argonaut on Thursday August 18, 2016 @03:03AM (#12310841)
> We badly need an alternative to Nvidia/CUDA for deep learning

The post I am replying to.

[ Parent | on HN ]