Christoph Ruegg

Yes, absolutely - now that we properly support .Net Standard, performance in general and leveraging SIMD in particular is indeed one of the main focus points.

For dense matrices this would directly map to [GEMM](https://github.com/mathnet/mathnet-numerics/blob/master/src/Numerics/Providers/LinearAlgebra/ILinearAlgebraProvider.cs#L231)?

Related: https://mathnetnumerics.codeplex.com/discussions/543114

Yes, I do plan to upgrade the MKL provider with the new MKL release. However, covering and exposing more areas (like the new DNN primitives) is another matter. For me...

- [x] Release MKL Providers v2.1.0 built against Intel MKL 2017: Windows - [ ] Release MKL Providers v2.1.0 built against Intel MKL 2017: Linux (without any code changes, just...

What would be typical criterion to control that iteration? absolute error threshold, error improvement threshold and iteration count?

Related: doesn't .NET 6 also introduce a shared Random instance? Maybe that could be useful for us also.

Note, the reason we pass in a seed is to avoid the problem Random used to have when you create multiple instances very quickly they are not guaranteed to be...

I have not merged yet because I first would like to bring out v5 which mainly changes a few things around native providers and how they are isolated, loaded an...

Just to clarify, this is about using the linear algebra providers directly from user code?