COMPMID-3160: Add Bfloat16 support in NEGEMMConvolutionLayer
Signed-off-by: Georgios Pinitas <georgios.pinitas@arm.com>
Change-Id: I0e449306c138a562ffc1455e76ec44b2fd059d85
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/2860
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Michele Di Giorgio <michele.digiorgio@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
diff --git a/src/core/NEON/kernels/arm_gemm/gemm_fp32.cpp b/src/core/NEON/kernels/arm_gemm/gemm_fp32.cpp
index 35493a6..96e3ce8 100644
--- a/src/core/NEON/kernels/arm_gemm/gemm_fp32.cpp
+++ b/src/core/NEON/kernels/arm_gemm/gemm_fp32.cpp
@@ -44,6 +44,7 @@
#include "kernels/sve_hybrid_fp32_mla_4VLx4.hpp"
#include "kernels/sve_interleaved_fp32_mla_3VLx8.hpp"
#include "kernels/sve_native_fp32_mla_4VLx4.hpp"
+#include "kernels/sve_smallK_hybrid_fp32_mla_1VLx8.hpp"
namespace arm_gemm {
@@ -73,7 +74,14 @@
},
#ifdef __ARM_FEATURE_SVE
-// SVE native / hybrid methods
+// SVE smallk / native / hybrid methods
+{
+ GemmMethod::GEMM_HYBRID,
+ "smallK_hybrid_fp32_mla_1VLx8",
+ [](const GemmArgs &args) { return (args._Ksize <= 24) && !args._trA && args._pretransposed_hint; },
+ nullptr,
+ [](const GemmArgs &args) { return new GemmHybrid<smallK_hybrid_fp32_mla_1VLx8, float, float>(args); }
+},
{
GemmMethod::GEMM_HYBRID,
"hybrid_fp32_mla_4VLx4",