Return to BSD News archive
Path: euryale.cc.adfa.oz.au!newshost.anu.edu.au!harbinger.cc.monash.edu.au!news.mel.connect.com.au!munnari.OZ.AU!news.ecn.uoknor.edu!news.cis.okstate.edu!news.ksu.ksu.edu!news.physics.uiowa.edu!math.ohio-state.edu!uwm.edu!vixen.cso.uiuc.edu!newsfeed.internetmci.com!info.ucla.edu!agate!reason.cdrom.com!usenet From: "Jordan K. Hubbard" <jkh@FreeBSD.org> Newsgroups: comp.unix.bsd.freebsd.misc Subject: Re: Pentium Pro Compatibility? Date: Sun, 14 Jan 1996 23:32:32 -0800 Organization: Walnut Creek CDROM Lines: 24 Message-ID: <30FA0310.794BDF32@FreeBSD.org> References: <4dc4al$btp@makai.maui.net> <DL75vA.MtG@ritz.mordor.com> NNTP-Posting-Host: time.cdrom.com Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Mailer: Mozilla 2.0b5 (X11; I; FreeBSD 2.1-STABLE i386) To: Chris Mauritz <ritz@ritz.mordor.com> Chris Mauritz wrote: > I was told that the problem was not with the cpu, but with the > new PCI chipset. It apparently introduces a serious bottleneck > with respect to cpu to main memory bandwidth. I don't have Not quite. The details are as follows: With the first generation of the Orion chipset (the one currently being sold), there was a bug in the PCI write buffer logic. Vendors such as ASUS work around the problem by turning the PCI write buffer off all the time, and this generally works OK until you really start hammering the bus. We found this out when we tried to run 100Mb ethernet and found that the performance was half what it was with the Triton chipset. If you're not really beating on the bus, e.g. you just have a PCI scsi controller and 10Mb ethernet card, you probably will never notice the problem. If you try to do more than this, you'll fall over kicking. We've "solved" the problem by getting a machine directly from Intel with the new revision of the Orion chipset, but it's not a for-sale item yet. :-( -- - Jordan Hubbard President, FreeBSD Project