{"type":"post","id":"flamia-flash-microscaling-attention-for-fast-and-causally-6740f5","slug":"flamia-flash-microscaling-attention-for-fast-and-causally-6740f5","title":"FLAMIA: FLAsh MIcroscaling Attention for Fast and Causally Safe Training on Blackwell","description":"FLAMIA is a family of attention kernels for NVIDIA Blackwell GPUs that combines MXFP8 and mixed NVFP4/FP8 training paths.","retrieval_nugget":"FLAMIA is a family of attention kernels for NVIDIA Blackwell GPUs that combines MXFP8 and mixed NVFP4/FP8 training paths.","published_at":"2026-11-03","updated_at":"2026-11-03","record_date":"2026-11-03","date_kind":"published_at","topics":["inference","models","compute"],"entities":["FLAMIA","NVIDIA Blackwell","NVIDIA B200","NVIDIA B300","Fedor Velikonivtsev","Mikhail Koltakov","Denis Kuznedelev"],"related_patterns":[],"novelty_label":"new_fact","what_changed":"FLAMIA is a new Blackwell attention-kernel paper, reporting MXFP8 speedups and training-stability results not covered by the archive.","prior_refs":[],"source_url":"https://arxiv.org/abs/2611.03344","source_urls":["https://arxiv.org/abs/2611.03344"],"schema_version":"newruntime-agent-readable-v0.2","stable_id":"post:flamia-flash-microscaling-attention-for-fast-and-causally-6740f5","status":"published","visuals":[],"editor":{"role":"editor","name":"Andrey Reshetnikov","profile":"https://newruntime.com/owner-profile.md","profile_json":"https://newruntime.com/owner-profile.json","same_as":["https://www.linkedin.com/in/reshetnikov1/","https://t.me/lapetuse"]},"routes":{"html":"https://newruntime.com/posts/flamia-flash-microscaling-attention-for-fast-and-causally-6740f5/","markdown":"https://newruntime.com/posts/flamia-flash-microscaling-attention-for-fast-and-causally-6740f5.md","json":"https://newruntime.com/posts/flamia-flash-microscaling-attention-for-fast-and-causally-6740f5.json"},"markdown_url":"https://newruntime.com/posts/flamia-flash-microscaling-attention-for-fast-and-causally-6740f5.md","next_reads":[{"type":"topic","path":"/topics/inference/","url":"https://newruntime.com/topics/inference/","title":"Inference","reason":"Explore the inference topic hub.","media_type":"text/html","markdown_url":"https://newruntime.com/topics/inference.md","json_url":"https://newruntime.com/topics/inference.json"},{"type":"topic","path":"/topics/models/","url":"https://newruntime.com/topics/models/","title":"Models","reason":"Explore the models topic hub.","media_type":"text/html","markdown_url":"https://newruntime.com/topics/models.md","json_url":"https://newruntime.com/topics/models.json"},{"type":"topic","path":"/topics/compute/","url":"https://newruntime.com/topics/compute/","title":"Compute","reason":"Explore the compute topic hub.","media_type":"text/html","markdown_url":"https://newruntime.com/topics/compute.md","json_url":"https://newruntime.com/topics/compute.json"},{"type":"collection","path":"/posts/","url":"https://newruntime.com/posts/","title":"Field Notes","reason":"Open the latest editorial analysis.","media_type":"text/html","json_url":"https://newruntime.com/posts.json"},{"type":"collection","path":"/sources/","url":"https://newruntime.com/sources/","title":"Sources","reason":"See the original sources this archive cites.","media_type":"text/html"}]}