Class NVCooperativeMatrix2

java.lang.Object
org.lwjgl.vulkan.NVCooperativeMatrix2

public class NVCooperativeMatrix2 extends Object
This extension adds several new features building on the cooperative matrix types added in VK_KHR_cooperative_matrix. The goal is to add and accelerate features beyond just simple GEMM kernels, including adding support for type/use conversions, reductions, per-element operations, and tensor addressing, and also to improve usability and out-of-the-box performance by adding support for more flexible matrix sizes, and workgroup scope matrices with compiler-managed staging through shared memory.

The new functionality is defined by the SPV_NV_tensor_addressing and SPV_NV_cooperative_matrix2 SPIR-V extensions and can be used with the GLSL_NV_cooperative_matrix2 GLSL extension.

This extension includes support for enumerating the matrix types and dimensions that are supported by the implementation, and which specific features are supported.

Name String
VK_NV_cooperative_matrix2
Extension Type
Device extension
Registered Extension Number
594
Revision
1
Extension and Version Dependencies
VK_KHR_cooperative_matrix
SPIR-V Dependencies
Contact
Extension Proposal
VK_NV_cooperative_matrix2
Other Extension Metadata
Last Modified Date
2024-08-01
Interactions and External Dependencies
Contributors
  • Jeff Bolz, NVIDIA
  • Karthik Vaidyanathan, NVIDIA