Slacker News Slacker News logo featuring a lazy sloth with a folded newspaper hat
  • top
  • new
  • show
  • ask
  • jobs
Library
← Back to context

Comment by qeternity

1 day ago

PyTorch is only part of it. There is still a huge amount of CUDA that isn’t just wrapped by PyTorch and isn’t easily portable.

3 comments

qeternity

Reply

svara  1 day ago

... but not in deep learning or am I missing something important here?

  • qeternity  21 hours ago

    Yes, absolutely in deep learning. Custom fused CUDA kernels everywhere.

    • Scene_Cast2  20 hours ago

      Yep. MoE, FlashAttention, or sparse retrieval architectures for example.

Slacker News

Product

  • API Reference
  • Hacker News RSS
  • Source on GitHub

Community

  • Support Ukraine
  • Equal Justice Initiative
  • GiveWell Charities