IronPython Performance: On Mono, on x64 and compared to C#

By Michael Foord -
Everyone loves benchmarks, even though they are usually misleading at best. Of course I'm no exception. Here are three snippets of news on IronPython performance that have surfaced over the last week or so.

IronPython 2.0.1 and 2.6 Beta 1 compared to CPython and Jython (posted to Twitpic by felipez) on Mono revision 136556:


Using the pystones benchmark IronPython on Mono clocks in faster than CPython or Jython. Nice.

One particular bugbear of IronPython performance has been startup time. On 64bit Windows the startup time goes from being merely bad to positively ridiculous. This is because the .NET 64bit JIT is slow to do its job and IronPython generates a lot of code needing to be JITed at startup. A lot of work has gone into IronPython 2.6 to improve startup time (by reducing the amount of rule generation done on startup), and this blog entry Dave Fugate explains something they have done specifically to address startup perf on 64bit Windows - and shows off the results:
Allow me to elaborate on this perf gain...Dino made a change recently such that ipy.exe is now strictly a 32-bit only assembly. That is, it gets executed as a 32-bit CLR process on both x86 and x64 operating systems. ipy64.exe, despite what its name might imply, is a platform agnostic assembly matching the old behavior of ipy.exe in the sense that it gets executed as a 32-bit process on 32-bit OSes and 64-bit on 64-bit OSes. At least part of the perf gain here comes from the fact that my simple Hello World program is most definitely not consuming anywhere near 4 gigs of memory, and overall memory consumption goes up for 64-bit processes => remember the size of pointers gets doubled. Just checked and the physical memory working set of ipy.exe was 31K compared to 44K under ipy64.exe for the Hello World program. This ~30% difference in memory consumption helps explain why we get better perf under ipy.exe.
The next blog entry is in Chinese and compares performance of a few operations in IronPython and C#. As you might expect C# is faster in general, but interestingly IronPython is actually faster than C# for reflection.

Comments

Post a Comment

Popular posts from this blog

Extending Abobe Flash Player and AIR with Python and Ruby

By Michael Foord -
An intriguing blog entry about an Adobe experimental project. It is a C/C++ cross- compiler for the flash player and AIR. This means that you can (will be able to?) cross-compile the Python / Ruby/ lots of other language runtimes and then run them in Flash. The blog entry contrasts this approach to the DLR which requires the re-implementation of dynamic languages - hence IronPython and IronRuby. The project is already beyond proof-of concept, they even have Quake III cross-compiled and running in Flash. Extending Adobe Flash Player and AIR with C/C++ Cross-Compiler Quake 3 has already been ported to managed code , but this is even more impressive. Whether the Python interpreter can be cross-compiled to run on top of another runtime (with completely different basic types), without changing the language semantics at all, has yet to be proved - but it is a very interesting project. (It also means that Python/Ruby will possibly be enabled only by additional optional downloads - but don'
Read more

Should Python Projects Support IronPython and Jython?

By Michael Foord -
The maintainers of the Spring Framework for Python ( Spring Python ) wonder how much of a priority it is to have compatibility with platforms other than CPython. In other words, should they support IronPython and Jython - and if so what should they do about their C extension dependencies? This question sparks a good discussion and is relevant to many other Python projects and applications: Need for supporting IronPython, Jython? " One of the topics on my mind recently, as we consider refactoring the core container, is how we ensure that when we introduce internal dependencies, such as possibly the Amara libraries for XML processing, what secondary constraints we might be introducing. " " For example, Amara represents a very pythonic and therefore simple to read and use approach to developing XML processing, such as our very own XML application configuration parsers. So from that perspective, things are very attractive. But then you ask the question, "can I run this
Read more

Further Adventures of the Debugger

By Michael Foord -
Harry Pierson (Microsoft IronPython PM) has been experimenting with writing an IronPython debugger ( ipydbg ) on top of the .NET MDbg APIs. When we last left him , he had implemented Just My Code (JMC) support allowing you to step into IronPython code. Since then he has made a lot of progress including a colourful REPL (interactive console) and solving some difficult problems around getting to real values behind boxed IronPython variables (needed for showing local variables). He has blogged extensively on what he has been up to (more recent posts last): Showing Source Code Colorful Console IronPython ConsoleColorMgr (the colorful console code pulled out into a separate IronPython module) A Little Hack...err...Cleanup Getting Local Variables Displaying Values Command Routing Getting Arguments REPL Console A lot of functionality is now present. In the command routing entry Harry outlines the seven commands supported by the debugger so far: Continue, Quit, Show Stack Trace, Show Locals,
Read more