[MLAS] Riscv64 rvv kernel optimizations #32540
Reduce once per row in the riscv64 softmax kernels
0ff52591
Carry an accumulator per output position in the riscv64 NCHWc kernels
0ef7fc0e
Choose the riscv64 NCHWc register group from the vector length
6925f2c3
Extend the riscv64 depthwise convolution to any kernel shape
2c04be81
Zestion
changed the title Riscv64 rvv kernel optimizations [MLAS] Riscv64 rvv kernel optimizations 60 days ago
Add an RVV activation routine on riscv64
b44bb448
Zestion
force pushed
from
7a9b090e
to
b44bb448
60 days ago
mirounga
approved these changes
on 2026-09-17
mirounga
merged
a76f0666
into main 53 days ago
Assignees
No one assigned
Login to write a write a comment.
Login via GitHub