From owner-freebsd-hardware Thu Sep 19 00:23:55 1996 Return-Path: owner-hardware Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id AAA16477 for hardware-outgoing; Thu, 19 Sep 1996 00:23:55 -0700 (PDT) Received: from mail.crl.com (mail.crl.com [165.113.1.22]) by freefall.freebsd.org (8.7.5/8.7.3) with SMTP id AAA16399; Thu, 19 Sep 1996 00:23:39 -0700 (PDT) Received: from GndRsh.aac.dev.com by mail.crl.com with SMTP id AA12790 (5.65c/IDA-1.5); Wed, 18 Sep 1996 21:29:10 -0700 Received: (from rgrimes@localhost) by GndRsh.aac.dev.com (8.7.5/8.7.3) id VAA05547; Wed, 18 Sep 1996 21:23:44 -0700 (PDT) From: "Rodney W. Grimes" Message-Id: <199609190423.VAA05547@GndRsh.aac.dev.com> Subject: Re: INN history file and disk I/O In-Reply-To: from Michael Beckmann at "Sep 19, 96 03:26:10 am" To: petzi@zit.th-darmstadt.de (Michael Beckmann) Date: Wed, 18 Sep 1996 21:23:44 -0700 (PDT) Cc: spork@super-g.com, isp@freebsd.org, hardware@freebsd.org X-Mailer: ELM [version 2.4ME+ PL25 (25)] 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 > Hi, > > > Buyer beware, Gigabyte is playing games with the specifications and > > register usage of the Triton-I and Triton-II chipsets to get away with > > these 6 SIMM slots. These chip sets have 8 programmable DRAM row size > > registers, each _SIDE_ of a SIMM requires one of them to be programmed, > > you can _ONLY_ run 4 double sided simms with these chip sets. > > According to the HX manual you can use 6 SIMMs with 8 MB each in it, and > these are double-sided. Then Gigabytes HX manual is in conflict with the Intel 82439HX databook, and somehow I trust the databook far more than I do Gigabyte MB manual. > OTOH, who cares whether one can put 6 double sided SIMMs in it. Anyone who is buying this board thinking they can load it with 6 32MB simms for a total memory capacity of 192MB, thats who cares!! > There are > so many valid combinations that I don't see this as a problem. Trust me, as a system builder, it _IS_ a problem. Both at initial machine configuration time (if I preload the system with 4 32MB simms there is no way to field expand the memory without tossing what is already in it away). Now your clone a day shop down the street wouldn't think twice about doing that because he usually doesn't really care about the customer 3 to 6 months down the road. I do care, and don't like to give customers false senses of upgradeabiliy. > The 586 ATE and HX mainboards I have used work just fine, I haven't > seen one fail yet. The 586ATE board does infact run just fine, and I can't recall exactly what it was about it that stopped me from adding it to my product line, perhaps that it offered nothing over my current product line. > That's just my personal experience. I'm not religious about mainboards, > but I don't have the impression that Asus boards have significantly fewer > bugs and problems. Gigabyte definitely belongs to the good equipment. I can agree that buying Gigabyte would be better than buying a lot of other boards out there. > > > Gigabyte dirty little move is to _ONLY_ allow single sided SIMMS in 2 > > Wow, you make it sound as if Gigabyte were posessed by the wicked one ;-) I consider miss leading datasheets to be wicked, yes I do consider Gigabytes marking department to be ``posessed by the wicked one'' for the particular way the use the 6 simm trick to fool people into thinking they can get more memory on there boards than any one elses. -- Rod Grimes rgrimes@gndrsh.aac.dev.com Accurate Automation Company Reliable computers for FreeBSD