VK_NV_cooperative_matrix_decode_vector(3) Manual Page

690

1

Not ratified

VK_NV_cooperative_matrix_decode_vector

This extension adds a device feature for the SPIR-V CooperativeMatrixDecodeVectorNV capability and the SPV_NV_cooperative_matrix_decode_vector SPIR-V extension.

When enabled, shaders may use an optional decode vector operand with OpCooperativeMatrixLoadTensorNV (see that SPIR-V extension) to load multiple block-adjacent elements per decode callback invocation.

An implementation must also support VkPhysicalDeviceCooperativeMatrix2FeaturesNV::cooperativeMatrixBlockLoads when exposing this extension.

No cross-references are available

For more information, see the Vulkan Specification.

This page is a generated document. Fixes and changes should be made to the generator scripts, not directly.

Copyright 2014-2026 The Khronos Group Inc.

SPDX-License-Identifier: CC-BY-4.0