Meta's 3.2K-line Triton JFA Kernel Beats FA4 on Blackwell
Meta's PyTorch team introduced Jagged Flash Attention (JFA), a 3.2K-line TLX-based attention kernel for Blackwell GPUs serving its GEM ads model, running 13% faster in forward and 50% faster in backward than FA4.
2026-10-02 ~ 2026-10-02 · 2 related posts
- PyTorch's TLX-based JFA kernel beats FlashAttention-4 by 13% fwd, 50% bwd on B200 — PyTorch · 2026-10-02
- Meta's TLX Jagged Flash Attention: 3.2K lines of Triton, beats FA4 by 13% forward / 50% backward — lmoroney · 2026-10-02