Whenever a native BLAS implementation isn't available on the system, Spark
automatically falls back onto a Java implementation. With the recent release
of the Vector API in the OpenJDK [1], we can use hardware acceleration for such
operations.

This patch introduces a VectorizedBLAS class which implements such
hardware-accelerated BLAS operations. This feature is hidden behind the
"vectorized" profile that you can enable by passing "-Pvectorized" to sbt or
maven.

The Vector API has been introduced in JDK 16. Following discussion on the
mailing list, this API is introduced transparently and needs to be enabled
explicitely.

[1] https://openjdk.java.net/jeps/338

All implmentations beat f2j on microbenchmarks on x86 (w/ AVX-2).

See https://github.com/luhenry/vectorizedblas/releases/tag/v0.1.5 for more details

It simplifies the build process, and will allow for faster iterations on my end

We still use it for the benchmarks but it should go away at some point

Brings acceleration for sscal, dgemm[N,N], dgemm[N,T], dgemv[N], sgemv[N] and sgemv[T]

This reverts commit bc9cfb0.

…lg.BLAS instead