Date: Thu, 18 Nov 2004 19:21:35 -0800 (PST) From: Arne "Wörner" <arne_woerner@yahoo.com> To: Luke <lukem@cse.unsw.EDU.AU> Cc: freebsd-performance@freebsd.org Subject: Re: I've ran out of ideas Message-ID: <20041119032135.89880.qmail@web41205.mail.yahoo.com> In-Reply-To: <Pine.LNX.4.61.0411191408440.11282@wagner.orchestra.cse.unsw.EDU.AU>
next in thread | previous in thread | raw e-mail | index | archive | help
--- Luke <lukem@cse.unsw.EDU.AU> wrote: > On Thu, 18 Nov 2004, Arne Wörner wrote: > > Hmm... Since /dev/zero delivers on my box about 100-300MByte/sec > > (100Mbyte/sec for 512byte block size), it looks like your network driver > > hinders the throughputting proceedings (I assume, that your other > > network devices can handle 90Mbit/sec; no ethernet collisions, no > > concurrency, ...). Are you sure, that Linux performs better in the same > > setting? > > I think you must mean Mbps, not MBps. If it were bytes, you would be > saturating 2.5 gigabit links with a single connection. Not to mention > most PCI busses. > neo# dd if=/dev/zero of=/dev/null bs=512k count=1000 1000+0 records in 1000+0 records out 524288000 bytes transferred in 1.475860 secs (355242396 bytes/sec) I meant /dev/zero, which does not use PCI or so (it just produces zeros; that's why it is so fast). Just the bus between processor and main memory, I think... With my 133MHz FSB(?) my DDRAM can do 8Gbit/sec??? I am not so good in such things... Looks like each byte gets copied three times before it reaches /dev/null... Does somebody agree? Maybe my contributions are too weird and misleading, but I do not see, that they are technically wrong. -Arne __________________________________ Do you Yahoo!? Meet the all-new My Yahoo! - Try it today! http://my.yahoo.com
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20041119032135.89880.qmail>