LAPACK, every mention
2 scenes · ← back to LAPACK
tap a year for its mentions
Verbatim, from the transcripts: the passages where LAPACK comes up
Multi-GPU Kernels, Intelligence per Watt, Heterogeneous Inference, and More | YC Paper Club · Y Combinator
- ▶ 41:47 unnamed speaker The API to use this in PyTorch is, is, is basically just calls LAPOC under the hood for CPU or Q solver for GPU.
- ▶ 44:04 unnamed speaker And the idea was, you know, at first it was sort of, like, vibe coded in the sense of, like, it was a POC that ported Torch to Python, so basically it just, like, went from Lua to PyTorch Ops, and you could, like, it was verifiable,…