PyPy on Pi
While we love all programming languages equally here at the Foundation, we do love Python an awful lot. Most users run their code under the “default” CPython interpreter, but over the last few years the PyPy project has made great strides in producing an highly compatible alternative interpreter with an integrated tracing JIT compiler. On x86 platforms this can improve the performance of some workloads by a factor of ten or more, and the PyPy team are now bringing the same sort of boost to the ARM world.
You can download an Pi-compatible alpha release of PyPy for ARM and see some benchmarks here. We’re proud to have been able to contribute a small amount of funding to the latter stages of this project; over the next few weeks we’ll be running an irregular series highlighting some of the other open source projects that we’ve been contributing to.
Daniel Radcliffe (Dan3008)
Wow. I use pypy on my laptop. Soooooo glad its coming to the pi :D
Great to see the RPF feeding some of its profits back into the open-source community :) Warm fuzzy wishes to you all.
We’re sponsoring quite a lot of open-source work at the moment; we’ll share what we’ve been up to when those projects are a bit further progressed. Exciting times!
Relay hopes this could speed Django up to a usable speed. Tried on my 256MB version and it varied between 2 times faster and ten times slower (memory usage was full). Would be interesting to see some benchmarks on the 512MB version.
i can now code in (almost) anything but now it’s our job to poo poo languages that don’t help new learners and good web standards
would love if pypy come to support RPi.GPIO
Comments are closed