Hacker News Reader

Agree it is a poor choice in name.

As for why CUDA, probably because that is what 90-something% of the ML community uses for this. Network affect.

Nothing wrong with the names, it's a name. Sure being "dark" might be a poor fit for code that's open. It's already has other meanings too, but in the end, it's name.

sigh people just can't think of a better brand name. Agree.

My bold guess is because there are more training machines run on Nviada GPUs as opposed to AMD's.

> My bold guess is because there are more training machines run on Nviada GPUs as opposed to AMD's.

Likely not so bold. AMD GPUs are good, but a quite a bit more pain to program compared to NVIDIA with CUDA (OpenCL on NVIDIA is more or less useless).

why CUDA, and not OpenCL?

The library appears to be using cuBLAS; what's the OpenCL alternative with comparable run-time performance and functionality?

Well there is http://developer.amd.com/tools-and-sdks/opencl-zone/acl-amd-...