Implement `MulAssign` (eager scaling) using SIMD

japaric-archived / linalg.rs

[INACTIVE]

Apache License 2.0

29 stars 1 forks source link

Implement `MulAssign` (eager scaling) using SIMD #20

Closed japaric closed 9 years ago

japaric commented 9 years ago

There is no BLAS routine for an eager mat *= scalar. f32x4/f64x2 can be used to provide explicit SIMD acceleration, though it will only work on continuous chunks of memory (Mat/Row<&[T]>/Col<&[T]>).

japaric commented 9 years ago

There is no BLAS routine for an eager mat *= scalar

Was wrong, there is a scal routine. This has been implemented in #51