From owner-freebsd-hardware Mon May 13 11:30:08 1996 Return-Path: owner-hardware Received: (from root@localhost) by freefall.freebsd.org (8.7.3/8.7.3) id LAA11362 for hardware-outgoing; Mon, 13 May 1996 11:30:08 -0700 (PDT) Received: from neon.Glock.COM (neon.glock.com [198.82.228.159]) by freefall.freebsd.org (8.7.3/8.7.3) with ESMTP id LAA11330 Mon, 13 May 1996 11:30:02 -0700 (PDT) Received: (from mmead@localhost) by neon.Glock.COM (8.7.5/8.7.3) id OAA02591; Mon, 13 May 1996 14:29:42 -0400 (EDT) From: "matthew c. mead" Message-Id: <199605131829.OAA02591@neon.Glock.COM> Subject: Re: Triton chipset with 256k cache caches 32M only? To: blh@nol.net (Brett L. Hawn) Date: Mon, 13 May 1996 14:29:42 -0400 (EDT) Cc: jgreco@brasil.moneng.mei.com, hackers@FreeBSD.ORG, hardware@FreeBSD.ORG In-Reply-To: from "Brett L. Hawn" at May 13, 96 01:24:52 pm X-Mailer: ELM [version 2.4 PL24 ME8a] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-hardware@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Brett L. Hawn writes: > On Mon, 13 May 1996, matthew c. mead wrote: > > Joe Greco writes: > > > > I've got two machine with moderately fast CPUs in > > > > them. One is a Cyrix 6x86 120+ (@100Mhz), and the other > > > > is a P90 (clocked to 100Mhz). When I have 40M in the > > > > machines, the upper 8M is not cached, and my performance > > > > is roughly 2/3 of that when they just have 32M and all of > > > > the memory is cached. > > > > Does anyone know for sure whether or not 256K cache > > > > Triton chipsets only cache up to 32M? Anyone know what > > > > I can do to get the other 8M cached as well? I'd really > > > > like to have that extra 8M in there, but at 2/3 the > > > > performance, it aint gonna happen. Any help is greatly > > > > appreciated! > > > My impression was 64M, based on past discussions with Rod > > > Grimes. > > Hmm. That would then imply that there's something wrong > > with each of these boards, or that the manufacturer is lazy. > > Does anyone know if going to 512K cache will allow me to > > cache on all 40M ram? Thanks! > Assuming that these are Triton-1 chipsets you will find that > anything over 64m leads to non-caching. I would highly suggest > getting some of the new ASUS (just my particular favorite) tr-2 > chipset motherboards, these solve the caching problem along > with many of the other inherent bugs of tr-1 chipsets. Well, I'm positive that anything over 64M will lead to non-caching. I'm also positive that anything over 32M leads to non-caching. Any ideas on what I should do to get the upper 8M (megs 32-39) cached? I'm thinking of purchasing a 512k COAST module, but I want to make sure that will do it before I buy it. -matt -- Matthew C. Mead mmead@Glock.COM http://www.Glock.COM/~mmead/