bohrium

Location: Denmark Remote: Yes, only remote Willing to relocate: No Technologies: Ruby, Ruby on Rails, React, PostgreSQL CV: https://mads.ohmbrix.dk/files/cv.pdf Email: REMOVED EDIT: Hi everybody. I have already gotten so many emails with job opportunties. I am so grateful for this community! Thanks a million. -- Hi HN. I'm Mads. I did my PhD working on Bohrium (https://github.com/bh107/bohrium). Bohrium is a framework that sits on top NumPy and intercepts all calls. When you have a side-effect in your Python code, Bohrium will take over and JIT-compile an OpenCL kernel, so your NumPy can run on your GPU. This is all done automatically, so all you need to change is `import numpy as np` to `import bohrium as np`. For Bohrium I also created a Ruby frontend, so you can use custom `BhArray` to JIT-compile OpenCL kernels from Ruby co

Veros is a port of a traditional, MPI-parallelized, Fortran ocean model to Python / NumPy. All parallelism is handled implicitly by Bohrium (https://github.com/bh107/bohrium), so all core routines just consist of basic NumPy-like array operations. The best part: Turns out we actually perform half as good as the Fortran reference on 24 CPUs, and as good on a GPU.

We do something similar with Bohrium[1]. Bohrium intercepts your NumPy calls and compile a kernel for either CPU, Multicore CPU, or GPU and run that instead. It will also fuse together all NumPy calls until one has a side-effect, that the run-time needs. We are currently building smart filters into it as well, so that we can do byte-code optimizations. [1] https://github.com/bh107/bohrium