matrixmultiply_mt 0.2.1

Multithreaded Fork of bluss's matrixmultiply crate. General matrix multiplication of f32 and f64 matrices in Rust. Supports matrices with general strides. Uses a microkernel strategy, so that the implementation is easy to parallelize and optimize. `RUSTFLAGS="-C target-cpu=native"` is your friend here.
Documentation 
1
2
3
4
5
6
7


test mat_mul_f32::mix10000         ... bench:  16,291,680 ns/iter (+/- 1,350,524)
test mat_mul_f64::mix10000         ... bench:  27,606,065 ns/iter (+/- 875,748)
test mat_mul_f32::m127             ... bench:     208,380 ns/iter (+/- 3,311)
test mat_mul_f64::m127             ... bench:     343,568 ns/iter (+/- 17,215)
test nonative_mat_mul_f32::m127             ... bench:     291,936 ns/iter (+/- 10,408)
test nonative_mat_mul_f64::m127             ... bench:     478,248 ns/iter (+/- 4,558)