I am aware about the interest in integrating Yeppp! into Julia (
see discussion here) and numexpr (
see this issue). Additionally, I plan to add Yeppp! function declarations for Cython soon. The repository also includes a sample of R bindings (
see here), albeit they are far from production quality. I do not have information about native (i.e. not via bindings) Yeppp! support in numpy, scipy, numba, or R, so it could be better to ask in those communities.
However, I don't think that Yeppp! could be a replacement for MKL. First, MKL has a lot more functionality (BLAS, FFT, numerical optimization) while Yeppp! currently focuses on vector operations. Secondly, MKL tries to be compatible with other popular interfaces (e.g. BLAS for Linear Algebra operations and FFTW for FFT) while Yeppp! uses its own naming conventions and interfaces for all modules.
I will be TAing CSE 6230 class (High-Performance Computing: Tools and Applications) for the online Masters in the Fall. Hope to see you there!
Regards,
Marat