Date: Sun, 3 Aug 1997 22:49:38 +0200 (MESZ) From: Robert Eckardt <roberte@MEP.Ruhr-Uni-Bochum.de> To: jmb@FreeBSD.ORG (Jonathan M. Bresler) Cc: tony@dell.com, cjs@portal.ca, freebsd@atipa.com, tom@sdf.com, hackers@FreeBSD.ORG Subject: Re: Pentium II? Message-ID: <199708032049.WAA06416@ghost.mep.ruhr-uni-bochum.de> In-Reply-To: <199708031631.JAA01116@hub.freebsd.org> from "Jonathan M. Bresler" at "3. Aug. 97 9:28:43"
next in thread | previous in thread | raw e-mail | index | archive | help
It was Jonathan M. Bresler who wrote: [...] > what we need is a benchmark that has a fixed data access pattern > and known data set size. better yet would be one that starts with > a very small data set and grows the data set till the computer starts > using disk. a graph of the results would show the speed of the > machine accross all its memory regimes. > > http://www.scl.ameslab.gov/scl/HINT/HINT.html What about lat_mem_rd from lmbench1.0 in the ports ? I would like to see some numbers for a Pentium II. >From a comparison between a PPro (thanks to Christoph Kukulies) and a Pentium-166, I learned that the L1 in the PPro seems to work with half the speed of the P5. lat_mem_rd 8 128 on a P5-166 $Id: lat_mem_rd.c,v 1.1 1994/11/18 08:49:48 lm Exp $ $Id: mhz.c,v 1.1 1994/11/18 08:51:55 lm Exp $ "stride=128 0.00049 6 0.02930 69 0.03125 69 2.00000 135 2.50000 137 > PPRO 200/256 > > $Id: lat_mem_rd.c,v 1.1 1994/11/18 08:49:48 lm Exp $ > $Id: mhz.c,v 1.1 1994/11/18 08:51:55 lm Exp $ > "stride=128 > 0.00049 10 <-- Das überrascht mich ! > 0.02930 30 > 0.03125 30 > 2.00000 196 > 2.50000 197 I noticed this speed difference in the run-time comparison of one of my numerical programs. Robert -- Robert Eckardt \\ FreeBSD -- solutions for a large universe.(tm) RobertE@MEP.Ruhr-Uni-Bochum.de \\ What do you want to boot tomorrow ?(tm) http://WWW.MEP.Ruhr-Uni-Bochum.de/~roberte For PGP-key finger roberte@gluon.MEP.Ruhr-Uni-Bochum.de
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199708032049.WAA06416>