https://github.com/Quafadas/vecxt

Aim

High performance shim to native vector libraries. Currently shims to BLAS.

Why?

To guarantee best in class performanc for vector operations in a cross platform context.

Status:

Need a reliable JS shim that imports via ES Modules.

Fundamentals work.

  • Vector
  • Matrix

Future

If we ever hit a serious performance bottleneck, consider a shim to https://developer.nvidia.com/gpu-accelerated-libraries

via slinc