Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Interesting that both de-novo and porting seems to have worked.

I do not understand why GGML is written this way, though. So much duplication, one variant per instruction set. Our Gemma.cpp only requires a single backend written using Highway's portable intrinsics, and last I checked for decode on SKX+Zen4, is also faster.



Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: