From owner-freebsd-smp Tue Sep 2 15:13:55 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id PAA01219 for smp-outgoing; Tue, 2 Sep 1997 15:13:55 -0700 (PDT) Received: from dyson.iquest.net (dyson.iquest.net [198.70.144.127]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id PAA01210 for ; Tue, 2 Sep 1997 15:13:49 -0700 (PDT) Received: (from root@localhost) by dyson.iquest.net (8.8.6/8.8.5) id RAA00487; Tue, 2 Sep 1997 17:13:39 -0500 (EST) From: "John S. Dyson" Message-Id: <199709022213.RAA00487@dyson.iquest.net> Subject: Re: SMP performance on P5 systems? In-Reply-To: <19970903005101.00664@hutcs.cs.hut.fi> from Antti-Pekka Liedes at "Sep 3, 97 00:51:01 am" To: apl@mail.cs.hut.fi (Antti-Pekka Liedes) Date: Tue, 2 Sep 1997 17:13:39 -0500 (EST) Cc: freebsd-smp@FreeBSD.ORG X-Mailer: ELM [version 2.4ME+ PL31 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-smp@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Antti-Pekka Liedes said: > > The current SMP performance on my Tyan Tomcat IIID dual pentium 200MMX > system is not quite what I had expected, as a simple benchmark, this is > what it takes for me to compile kernel, with UP kernel: > make -j3 283.19s user 32.56s system 94% cpu 5:32.83 total > and with SMP: > make -j3 341.23s user 134.72s system 183% cpu 4:18.80 total > > I was expecting the wall clock time to be less for the SMP kernel, also, > there seems to be quite a lot of system overhead :-/. Of course this > might be because of my relatively slow disks (I use striping on two > Quantum Fireballs, both on different channel of my AHA-3940U). The > benchmarks at www.freebsd.org SMP pages show a lot better performance. Is > it just that ppro beats pentium so much on SMP or what? > I don't know if this is why PPros might sometimes appear to be faster in SMP than P5, but each CPU having a 2nd level cache probably has something to do with it. -- John dyson@freebsd.org jdyson@nc.com