Date: Tue, 5 Oct 1999 09:33:07 -0600 From: Nate Williams <nate@mt.sri.com> To: SHUDO Kazuyuki <shudoh@muraoka.info.waseda.ac.jp> Cc: freebsd-java@FreeBSD.ORG Subject: Re: Announce: New release of JDK1.1.8 for FreeBSD Message-ID: <199910051533.JAA19125@mt.sri.com> In-Reply-To: <199910050635.PAA14746@cafe.muraoka.info.waseda.ac.jp> References: <199910041552.JAA14040@mt.sri.com> <199910050635.PAA14746@cafe.muraoka.info.waseda.ac.jp>
next in thread | previous in thread | raw e-mail | index | archive | help
> I guess that Volano mark requires not only performance > of bytecode execution, but also effective multithreading. Actually, in my own benchmarks making *HEAVY* use of sockets and such, I compared the Solaris green threads implementation (not kernel threads) vs. the static executable created by TowerJ and the difference was ~25%. At the time, TowerJ on Solaris was the *fastest* VM around. > There are works remains in the area of OS, yes it's FreeBSD. :) I'm less convinced of that than you, especially in extremely I/O bound applications. If you're I/O bound and use lots of threads, then it turns out that the green threads switching is actually *faster* because it doesn't have to do complete context switching of kernel threads. And, because I/O is 'mostly' handled by select inside the green threads code, it actually works pretty well for most things... Anyway, that has been my experience. I can't compare the FreeBSD stuff, because we have nothing like TowerJ. Nate To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-java" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199910051533.JAA19125>