Merge requests · Kleidi / KleidiAI

Merge trains

Add SME F32 GEMV pairing for the GEMM

!160 · created Oct 25, 2024 by Jakub Sujak
feature
- Merged
- +2
- 3
- Approved
updated Oct 31, 2024
Add SME Int8 GEMM

!156 · created Oct 22, 2024 by Suhail M 0.5.0 release
feature
- Merged
- +2
- 101
- 1
- Approved
updated Nov 29, 2024
Add bf16 Lhs/Rhs packed gemv kernel

!154 · created Oct 21, 2024 by Gunes Bayir
feature
- Merged
- +5
- 75
- Approved
updated Oct 30, 2024
Optimize F32 <- QAI8DXP (LHS) x QSI4CXP (RHS) for SME

!153 · created Oct 21, 2024 by Anitha Raj Next release
feature
- Merged
- 69
- Approved
updated Dec 24, 2024
Add RHS packing kernel for SME2 Int8 GEMM

!152 · created Oct 18, 2024 by Viet-Hoa Do 0.5.0 release
feature
- Closed
- 5
- 0 of 1 Approvals
updated Nov 29, 2024
Add transposed RHS packing for SME2 FP32 GEMM

!150 · created Oct 16, 2024 by Viet-Hoa Do
feature
- Merged
- +3
- 20
- Approved
updated Oct 29, 2024
Add SME F32 GEMV micro-kernel

!137 · created Oct 01, 2024 by Jakub Sujak
feature
- Merged
- +2
- 14
- Approved
updated Oct 16, 2024
Draft: Add bf16 hybrid gemm kernels

!135 · created Sep 30, 2024 by Gunes Bayir
feature
- Closed
- 3
- 0 of 1 Approvals
updated Jan 10, 2025
Add fp16 hybrid gemm kernel with 6x32 output block size

!134 · created Sep 30, 2024 by Gunes Bayir
feature
- 3
- 0 of 1 Approvals
updated Nov 04, 2024
Add bf16 Lhs/Rhs packed gemm kernels and packing functions

!133 · created Sep 30, 2024 by Gunes Bayir
feature
- Merged
- +5
- 157
- Approved
updated Oct 31, 2024
Add FP32 SME2-kernel to the benchmark suite.

!115 · created Sep 18, 2024 by Jens Elofsson
feature
- Merged
- +3
- 17
- Approved
updated Oct 09, 2024
Add FP32 GEMV micro kernel

!87 · created Aug 14, 2024 by Felix Johnny Thomasmathibalan
feature
- Merged
- 5
- Approved
updated Aug 30, 2024
Matmul int4 micro-kernels for QAI8DX (LHS) x QSI4C32 (RHS) -> F32

!28 · created Jun 10, 2024 by Gian Marco Iodice
feature
- Merged
- +3
- 79
- 2
- Approved
updated Sep 27, 2024