in reply to Re: autobox performance:a real-world comparison
in thread autobox performance:a real-world comparison
The cause is memory growth! Despite that they both process identical datsets:
I'd be interested to hear if the autobox also grabs a heap of memory on your systems?
I also instrumenting them both with NYTProf, but the output is so bewildering (and large) that I'm not able to make any sense of it. If there is anyone out there that wants to try and fix this, I'll make the NYTProf output, plus any other info I can supply, to them.
|
|---|