GPU Infrastructure Inference Optimization

FlashQLA

High-performance linear attention kernel library built on TileLang.

Open-source high-performance linear attention kernel library built on TileLang for GDN Chunked Prefill, with fused forward/backward optimizations and NVIDIA Hopper performance tuning.

Recent stories

1 linked story

releasePRIMARY2026-04-29

FlashQLA releases TileLang linear-attention kernels with 2–3x forward speedups

Alibaba Qwen introduced FlashQLA, a TileLang-based linear-attention kernel stack that reports 2–3x faster forward passes and 2x faster backward passes. The release gives edge and long-context deployments a new optimization lever below the model layer itself.