Hacker News Reader

This is great, but it needs a corresponding memory usage graph.

What is the best way to make use of PyPy to speed up applications? Every time I hear about PyPy, it's always in the context of performance, but these stats indicate that it's actually slower than CPython at pretty much everything -- not just the interpreter, but also when using the JIT.

So my question is: how are you supposed to make use of PyPy when you have a Python script that demands performance?

I'm not sure why you think pypy is slower than cpython "at pretty much everything" after looking at this site. Out of the 21 benchmarks given, 10 are 2x or more faster than cpython, 10 are between 2x faster and 2x slower and only 1 is more than 2x slower.

The go programming language also has such a dashboard. http://godashboard.appspot.com/benchmarks Takes a while to load, fyi. I think dashboards like these are a great idea and all implementations should have something like it.

I think Miquel Torres did an excellent job putting this site together. For anyone that is interested the code for the site can be found at http://github.com/tobami/codespeed

Did anyone find the the sources for the benchmarks?

Benchmark sources are here: http://codespeak.net/svn/pypy/benchmarks/

PyPy Speed Center