● Open SourceNov 7, 2026 · 00:281 source

Zyphra ZYFA: FP8 FlashAttention for AMD MI355X, Up to 1.76x Faster

From marktechpost.com

AI;DR

Zyphra released ZYFA, open-source FP8 and BF16 prefill attention kernels for AMD Instinct MI355X. On real model activations, its FP8 kernels run 1.16-1.76x faster than AMD’s AITER with 62-78% lower error at head dimension 128.

Read originalSign in to suggest

Sources

#inference#open-sourceMarkdownScore 0.26.3k tokens
Image: marktechpost.com
Read original