search

ACEsuit / Polynomials4ML.jl

Polynomials for ML: fast evaluation, batching, differentiation
MIT License
12 stars 5 forks source link

Batching is sometimes slower #7

Closed cortner closed 2 years ago

cortner commented 2 years ago

In the current implementation of RYlmBasis, the batched version is slower than the serial version + doesn't allow us to @avx. it is completely unclear why?

Should make a general performance study for all the bases implemented here.

cortner commented 2 years ago

this was due to a performance bug in the code - close for now until it comes up again.