{"version":"1.0","type":"card","id":"6467c56f-01dd-44fd-8dc7-de4f3aaf8448","url":"https://stacklist.com/card/6467c56f-01dd-44fd-8dc7-de4f3aaf8448","title":"If you're still training in FP16, you're leaving half your GPU on the table.","source_url":"https://www.linkedin.com/posts/paoloperrone_if-youre-still-training-in-fp16-youre-share-7480370913650176000-CIDz/?utm_source=share&utm_medium=member_ios&rcm=ACoAAAI21ZsBNnZPaKuTab7nquKLCveUW7o-1DE","note":"This page discusses the advantages of using 8-bit floating point formats over FP16 for training neural networks. It highlights the benefits of E4M3 and E5M2 formats for different training phases, emphasizing memory savings, increased throughput, and maintaining accuracy.","image":{"url":"https://ucarecdn.com/b91bb2ae-6337-41c4-9030-add02e079f46/","alt":"If you're still training in FP16, you're leaving half your GPU on the table.","width":1200,"height":750},"stack":{"id":"ed01fcda-d7b5-4a4b-bdfa-bc07c9df5228","title":"Local AI & GPUs","url":"https://stacklist.com/c/technology/stack/ed01fcda-d7b5-4a4b-bdfa-bc07c9df5228"},"created_at":"2026-07-29T02:39:13.556Z","updated_at":null,"aco":null,"_links":{"self":"/api/public/card/6467c56f-01dd-44fd-8dc7-de4f3aaf8448.json","html":"https://stacklist.com/card/6467c56f-01dd-44fd-8dc7-de4f3aaf8448","md":"/api/public/card/6467c56f-01dd-44fd-8dc7-de4f3aaf8448.md","stack_json":"/api/public/stack/ed01fcda-d7b5-4a4b-bdfa-bc07c9df5228.json"}}