commit | c952596e70f2fe0073029f053e329a4e930ced8c | [log] [tgz] |
---|---|---|
author | Ramy Elgammal <ramy.elgammal@arm.com> | Fri May 19 14:23:37 2023 +0100 |
committer | Ramy Elgammal <ramy.elgammal@arm.com> | Fri Jun 23 20:06:45 2023 +0000 |
tree | 1cf9b1c87c2288d6af436b570802d9cc6e8b30b5 | |
parent | 47a50ef12f513cfa8fde6673b8a61ed0f2d0fbaa [diff] [blame] |
Implement FP32/FP16 MatMul NT/T kernel using the MMUL extension Resolves COMPMID-6195 Signed-off-by: ramy.elgammal@arm.com <ramy.elgammal@arm.com> Change-Id: I8e85fe73308ed84ebb142d6d6d1562b62dddfaa5 Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/9819 Reviewed-by: SiCong Li <sicong.li@arm.com> Benchmark: Arm Jenkins <bsgcomp@arm.com> Tested-by: Arm Jenkins <bsgcomp@arm.com> Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
diff --git a/src/gpu/cl/kernels/ClMatMulNativeMMULKernel.h b/src/gpu/cl/kernels/ClMatMulNativeMMULKernel.h index 26fe08c..79f675d 100644 --- a/src/gpu/cl/kernels/ClMatMulNativeMMULKernel.h +++ b/src/gpu/cl/kernels/ClMatMulNativeMMULKernel.h
@@ -86,6 +86,7 @@ private: int _m{ 1 }; int _n{ 1 }; + int _k{ 1 }; }; } // namespace kernels } // namespace opencl