{"id":12369,"date":"2025-09-11T10:00:34","date_gmt":"2025-09-11T17:00:34","guid":{"rendered":"https:\/\/devblogs.microsoft.com\/directx\/?p=12369"},"modified":"2025-09-15T16:25:11","modified_gmt":"2025-09-15T23:25:11","slug":"shader-model-6-9-and-the-future-of-cooperative-vector","status":"publish","type":"post","link":"https:\/\/devblogs.microsoft.com\/directx\/shader-model-6-9-and-the-future-of-cooperative-vector\/","title":{"rendered":"Shader Model 6.9 and The Future of Cooperative Vector"},"content":{"rendered":"

D<\/span>irect<\/span>3D<\/span>\u00a0<\/span>Cooperative Vector\u00a0<\/span>is Evolving<\/span><\/span><\/h2>\n

DirectX continues to evolve to support the growing demands of AI and machine learning workloads. As part of this effort, we plan to introduce an expanded suite of linear algebra capabilities in a future Shader Model release, unlocking even more robust GPU acceleration for bedrock AI\/ML computations within the graphics pipeline.<\/span>\u00a0<\/span><\/p>\n

This work\u00a0extends\u00a0our\u00a0experimental\u00a0<\/span>Cooperative Vector<\/span><\/b> implementation by introducing accelerated matrix-matrix operations, crucial for scaling AI\/ML workloads beyond what lightweight vector-matrix alone can handle. This will unlock new performance gains in compute shaders by enabling efficient, large-block tensor operations, making it ideal for demanding tasks like super resolution, LLM inference, and computer vision. Both matrix-matrix and vector-matrix capabilities will be unified under a single design, providing a consistent and scalable foundation for high-performance shader code for AI\/ML scenarios.\u00a0<\/span>\u00a0<\/span><\/p>\n

Developer feedback has been central to this evolution,\u00a0from early experimentation to shaping the unified design. If you have\u00a0additional\u00a0input,\u00a0we\u2019d\u00a0love to hear it \u2013\u00a0contributions to the spec\u00a0can be made\u00a0<\/span>here<\/span><\/a>!<\/span><\/p>\n

Development Notes<\/span><\/span><\/h4>\n