There’s a readme with a lot of output dumped on it here: https://github.com/chriselrod/TriangularMatrices.jl But the highlights are that, I found an fairly optimal “compute kernel” (at least for my architecture on Julia 0.7; performance is poor on Julia 0.6): julia> using StaticArrays, Triangula…

We can write an optimized BLAS library in pure Julia (please skip OP and jump to post 4)

simonbyrne June 13, 2018, 3:59am 2

That looks exciting! See also this earlier thread.

3 Likes

Topic		Replies	Views
[ANN]: PaddedMatrices.jl, Julia BLAS and partially sized arrays Package Announcements performance , blas	32	5857	July 5, 2020
@inbounds: is the compiler now so smart that this is no longer necessary? Performance	33	2965	July 16, 2018
Alternate BLAS libraries? General Usage blas	22	2990	July 4, 2020
Why mul! is so fast? General Usage question , linearalgebra	7	6921	November 26, 2019
Native Julia gemm implementation Performance	16	3613	May 3, 2018