[pypy-svn] r68017 - pypy/extradoc/talk/pycon2010
pedronis at codespeak.net
pedronis at codespeak.net
Wed Sep 30 12:09:16 CEST 2009
Author: pedronis
Date: Wed Sep 30 12:09:15 2009
New Revision: 68017
Modified:
pypy/extradoc/talk/pycon2010/jit_abstract.txt
Log:
be a bit less over-concrete
Modified: pypy/extradoc/talk/pycon2010/jit_abstract.txt
==============================================================================
--- pypy/extradoc/talk/pycon2010/jit_abstract.txt (original)
+++ pypy/extradoc/talk/pycon2010/jit_abstract.txt Wed Sep 30 12:09:15 2009
@@ -21,19 +21,14 @@
Proposed talk agenda:
-* Show various benchmarking data and compare different
- python interpreters (CPython, PyPy, Jython, IronPython, unladden swallow)
+* Show various benchmarking data and compare different Python interpreters
-* A bit of introduction into how PyPy's JIT work (it's tracing JIT,
- like tracemonkey JS interpreter, unlike Psyco or JVM).
+* A bit of introduction into how PyPy's JIT work (it's a generated tracing JIT,
+ like the tracemonkey JS interpreter, unlike Psyco or JVM JITs).
-* What sorts of programs JITs in general can speed up. What sorts of
- programs PyPy's JIT is better at. What similiar python constructs
- can yield different performance characteristics while JITting.
+* Tracing JITs focus on often executed loops. Which kind of programs
+ will the PyPy JIT speed-up , which constructs we know it is currently poor at.
* Future plans for PyPy, especially JIT-wise.
-The main wieght would be put on the point 3 (how can I write my program
-so JIT will make it fast).
-
.. _`faster than Psyco`: http://morepypy.blogspot.com/2009/09/first-results-of-jit.html
More information about the Pypy-commit
mailing list