diff options
| author | snadampal <87143774+snadampal@users.noreply.github.com> | 2024-02-11 07:22:33 -0600 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2024-02-11 15:22:33 +0200 |
| commit | a07d0fee1f05c5c1dc49948ae1a3293db017275f (patch) | |
| tree | 06614ff1364269493e4853333ced56802abd7284 /common/sampling.cpp | |
| parent | e4640d8fdf56f14a6db3d092bcd3d2d315cb5d04 (diff) | |
ggml : add mmla kernels for quantized GEMM (#4966)
* ggml: aarch64: implement smmla kernel for q8_0_q8_0 quantized gemm
armv8.2-a and above supports MMLA instructions that have higher
throughput than DOT. this commit adds mmla kernel for
q8_0_q8_0 gemm. The feature is enabled if the platform supports
"__ARM_FEATURE_MATMUL_INT8"
On AWS Graviton3 processors this kernel resulted up to 1.5x
improvement for prompt evaluation throughput compared to the
default sdot kernel.
* ggml: aarch64: implement smmla kernel for q4_0_q8_0 quantized gemm
armv8.2-a and above supports MMLA instructions that have higher
throughput than DOT. this commit adds mmla kernel for
q4_0_q8_0 gemm. The feature is enabled if the platform supports
"__ARM_FEATURE_MATMUL_INT8"
On AWS Graviton3 processors this kernel resulted up to 1.5x
improvement for prompt evaluation throughput compared to the
default sdot kernel.
* ggml: aarch64: implement smmla kernel for q4_1_q8_1 quantized gemm
armv8.2-a and above supports MMLA instructions that have higher
throughput than DOT. this commit adds mmla kernel for
q4_1_q8_1 gemm. The feature is enabled if the platform supports
"__ARM_FEATURE_MATMUL_INT8"
On AWS Graviton3 processors this kernel resulted up to 1.5x
improvement for prompt evaluation throughput compared to the
default sdot kernel.
* ggml: update unit tests for the new vec_dot interface
* llama.cpp: add MATMUL_INT8 capability to system_info
Diffstat (limited to 'common/sampling.cpp')
0 files changed, 0 insertions, 0 deletions