tag:blogger.com,1999:blog-3971202189709462152.post9211261260383281459..comments2024-03-11T12:50:02.036+01:00Comments on PyPy Status Blog: Some benchmarkingCarl Friedrich Bolz-Tereickhttp://www.blogger.com/profile/00518922641059511014noreply@blogger.comBlogger32125tag:blogger.com,1999:blog-3971202189709462152.post-17200233285616905632010-02-18T08:05:53.405+01:002010-02-18T08:05:53.405+01:00I'm curious why mandelbrot is much less accele...I'm curious why mandelbrot is much less accelerated than, say, nbody. Does PyPy not JIT complex numbers properly yet?Term Paperhttp://www.flashpapers.comnoreply@blogger.comtag:blogger.com,1999:blog-3971202189709462152.post-44143460368653874282009-12-01T20:07:52.508+01:002009-12-01T20:07:52.508+01:00I checked http://www.looking-glass.us/~chshrcat/py...I checked http://www.looking-glass.us/~chshrcat/python-benchmarks/results.txt but it doesn't have the data for unladen swallow. Where are the number?Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-3971202189709462152.post-17984830909274577752009-11-30T08:26:00.696+01:002009-11-30T08:26:00.696+01:00glad to see someone did something with my language...glad to see someone did something with my language shootout benchmark comment ;)¬¬https://www.blogger.com/profile/09229993367715169480noreply@blogger.comtag:blogger.com,1999:blog-3971202189709462152.post-78024241373532805212009-11-27T14:54:32.648+01:002009-11-27T14:54:32.648+01:00Do any of those benchmarks work with shedskin?Do any of those benchmarks work with shedskin?Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-3971202189709462152.post-12920647575612547292009-11-23T22:10:51.144+01:002009-11-23T22:10:51.144+01:00Tom is right, normalizing the graphs to cpython = ...Tom is right, normalizing the graphs to cpython = 1.0 would make them much more readable.<br />Anyway, this is a very good Job from Unhelpful.<br />Thanks!Luishttps://www.blogger.com/profile/01147433030878927988noreply@blogger.comtag:blogger.com,1999:blog-3971202189709462152.post-65863007509054273162009-11-23T20:24:13.029+01:002009-11-23T20:24:13.029+01:00it's called "The Computer Language
Benc...it's called "The Computer Language <br />Benchmarks Game" these days...Unknownhttps://www.blogger.com/profile/11240181077733786807noreply@blogger.comtag:blogger.com,1999:blog-3971202189709462152.post-9806979270613329802009-11-22T13:08:30.962+01:002009-11-22T13:08:30.962+01:00Nice comparisons - and micro-performance looking g...Nice comparisons - and micro-performance looking good. Congratulations.<br /><br />HOWEVER - there is no value in having three columns for each benchmark. The overall time is arbitrary, all that matters is relative so you might as well normalise all graphs to CPython = 1.0, for example. The relevant informtion is then easier to see!Tom Clarkehttps://www.blogger.com/profile/08786618190262592330noreply@blogger.comtag:blogger.com,1999:blog-3971202189709462152.post-39574028033512890292009-11-22T11:13:53.194+01:002009-11-22T11:13:53.194+01:00Would perhaps also be nice to compare the performa...Would perhaps also be nice to compare the performance with one the current Javascript-Engines(V8, SquirrelFish etc.)Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-3971202189709462152.post-50207043101388070192009-11-22T02:45:56.106+01:002009-11-22T02:45:56.106+01:00@Leo it's likely to take tons of memory, thoug...@Leo it's likely to take tons of memory, though.Benjamin Petersonhttps://www.blogger.com/profile/06955536323236904839noreply@blogger.comtag:blogger.com,1999:blog-3971202189709462152.post-41800609838021283812009-11-21T20:35:17.988+01:002009-11-21T20:35:17.988+01:00@Maciej Great! It'll be awesome to have a (ho...@Maciej Great! It'll be awesome to have a (hopefully much faster??) JITted build ... it currently takes my computer more than an hour ...Leonoreply@blogger.comtag:blogger.com,1999:blog-3971202189709462152.post-41860361796422309352009-11-21T18:47:42.054+01:002009-11-21T18:47:42.054+01:00antonio: good point. On the second thought, thoug...antonio: good point. On the second thought, though, it's not a *really* good point because we don't have _immutable_=True on floats either...Armin Rigohttps://www.blogger.com/profile/06300515270104686574noreply@blogger.comtag:blogger.com,1999:blog-3971202189709462152.post-57779595430481487902009-11-21T12:26:54.387+01:002009-11-21T12:26:54.387+01:00@Leo:
yes, we know that bug. Armin is fixing it r...@Leo:<br /><br />yes, we know that bug. Armin is fixing it right now on faster-raise branch.Maciej Fijalkowskihttps://www.blogger.com/profile/11410841070239382771noreply@blogger.comtag:blogger.com,1999:blog-3971202189709462152.post-47293350508071157492009-11-21T10:06:47.919+01:002009-11-21T10:06:47.919+01:00So, um, has anyone managed to get JIT-ed pypy to c...So, um, has anyone managed to get JIT-ed pypy to compile itself? <br /><br />When I tried to do this today, I got this:<br /><br />http://paste.pocoo.org/show/151829/Leonoreply@blogger.comtag:blogger.com,1999:blog-3971202189709462152.post-91306624409576034442009-11-21T10:05:17.793+01:002009-11-21T10:05:17.793+01:00This comment has been removed by the author.Antiplutocrathttps://www.blogger.com/profile/00307747349924930767noreply@blogger.comtag:blogger.com,1999:blog-3971202189709462152.post-15406828071211007052009-11-21T00:58:19.005+01:002009-11-21T00:58:19.005+01:00guys, sorry, who cares about *seconds*??
why didn...guys, sorry, who cares about *seconds*??<br /><br />why didn't you normalize to the test winners? :)vakhttps://www.blogger.com/profile/13157594553222689990noreply@blogger.comtag:blogger.com,1999:blog-3971202189709462152.post-47561232478137030312009-11-21T00:54:22.568+01:002009-11-21T00:54:22.568+01:00@Anon, @Benjamin
I've just noticed that W_Comp...@Anon, @Benjamin<br />I've just noticed that W_ComplexObject in objspace/std/complexobject.py is not marked as _immutable_=True (as it is e.g. W_IntObject), so it is totally possible that the JIT is not able to optimize math with complexes as it does with ints and floats. We should look into it, it is probably easy to discoverAntonio Cunihttps://www.blogger.com/profile/17017456817083804792noreply@blogger.comtag:blogger.com,1999:blog-3971202189709462152.post-91362826716909926452009-11-20T23:41:34.545+01:002009-11-20T23:41:34.545+01:00@Anon I'm only guessing. Our math is awfully f...@Anon I'm only guessing. Our math is awfully fast.Benjamin Petersonhttps://www.blogger.com/profile/06955536323236904839noreply@blogger.comtag:blogger.com,1999:blog-3971202189709462152.post-18752553227521196072009-11-20T16:35:26.999+01:002009-11-20T16:35:26.999+01:00Benjamin, is it really an issue with array? The in...Benjamin, is it really an issue with array? The inner loop just does complex arithmetic. --AnonAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-3971202189709462152.post-705081297670529432009-11-20T11:41:32.531+01:002009-11-20T11:41:32.531+01:00@holger: yes, that explains why regexps are not fa...@holger: yes, that explains why regexps are not faster in PyPy, but not why they are 5x or 10x slower. Of course our regexp engine is terribly bad. We should have at least a performance similar to CPython.Armin Rigohttps://www.blogger.com/profile/06300515270104686574noreply@blogger.comtag:blogger.com,1999:blog-3971202189709462152.post-46214412352688404282009-11-20T08:37:01.563+01:002009-11-20T08:37:01.563+01:00ASFAIU it's not PyPy's regex engine being ...ASFAIU it's not PyPy's regex engine being "bad" but rather the fact that the JIT generator cannot consider and optimize the loop in the regex engine, as it is a nested loop (the outer one being the bytecode interpretation one).holger krekelhttps://www.blogger.com/profile/00985924698593515074noreply@blogger.comtag:blogger.com,1999:blog-3971202189709462152.post-16180754190359686882009-11-20T08:11:37.309+01:002009-11-20T08:11:37.309+01:00How long until I can do
pypy-c-jit translate.py ...How long until I can do <br /><br />pypy-c-jit translate.py -Ojit targetpypystandalone.py<br /><br />?<br /><br />So far, when I try, I get<br /><br />NameError: global name 'W_NoneObject' is not defined<br />http://paste.pocoo.org/show/151829/Leonoreply@blogger.comtag:blogger.com,1999:blog-3971202189709462152.post-319567231306341842009-11-20T04:11:35.263+01:002009-11-20T04:11:35.263+01:00@Anon Our array module is in pure Python and much ...@Anon Our array module is in pure Python and much less optimized than CPython's.Benjamin Petersonhttps://www.blogger.com/profile/06955536323236904839noreply@blogger.comtag:blogger.com,1999:blog-3971202189709462152.post-72920284242939752222009-11-20T04:03:58.531+01:002009-11-20T04:03:58.531+01:00@wilk ./translate.py -Ojit targetpypystandalone.py...@wilk ./translate.py -Ojit targetpypystandalone.pyBenjamin Petersonhttps://www.blogger.com/profile/06955536323236904839noreply@blogger.comtag:blogger.com,1999:blog-3971202189709462152.post-38657337156889138462009-11-20T00:38:06.359+01:002009-11-20T00:38:06.359+01:00I'm curious why mandelbrot is much less accele...I'm curious why mandelbrot is much less accelerated than, say, nbody. Does PyPy not JIT complex numbers properly yet?Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-3971202189709462152.post-84291474919193797552009-11-19T17:04:05.799+01:002009-11-19T17:04:05.799+01:00Congratulations !
Please could you remember us ho...Congratulations !<br /><br />Please could you remember us how to build and test pypy-jit ?wilkhttps://www.blogger.com/profile/15960294417115031870noreply@blogger.com