From owner-freebsd-hardware Fri Oct 11 10:04:25 1996 Return-Path: owner-hardware Received: (from root@localhost) by freefall.freebsd.org (8.7.5/8.7.3) id KAA12161 for hardware-outgoing; Fri, 11 Oct 1996 10:04:25 -0700 (PDT) Received: from cs.rice.edu (cs.rice.edu [128.42.1.30]) by freefall.freebsd.org (8.7.5/8.7.3) with ESMTP id KAA12156 for ; Fri, 11 Oct 1996 10:04:22 -0700 (PDT) Received: from una.cs.rice.edu (una.cs.rice.edu [128.42.1.160]) by cs.rice.edu (8.7.1/8.7.1) with ESMTP id MAA22149 for ; Fri, 11 Oct 1996 12:04:09 -0500 (CDT) From: Robert J Fowler Received: (from rjf@localhost) by una.cs.rice.edu (8.7.3/8.7.3) id MAA20346 for hardware@freebsd.org; Fri, 11 Oct 1996 12:04:08 -0500 (CDT) Message-Id: <199610111704.MAA20346@una.cs.rice.edu> Subject: AMD 586 runs FreeBSD just FINE (fwd) To: hardware@freebsd.org Date: Fri, 11 Oct 1996 12:04:07 -0500 (CDT) Organization: CRPC / Dept. of Computer Science, Rice University Addr(Post): Rice U, MS-41, CRPC/CITI, P.O. Box 1892, Houston, Texas 77251 USA Addr(Express): CRPC/CITI, Rice U., MS-41, 6100 S. Main, Houston TX 77005 Phone: (713) 285-5176 X-Mailer: ELM [version 2.4 PL25] 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 Matthew Gessner announced -- > > Hello, all, > > A little status report for y'all. > > I just bought one of those them there fancy shmancy CPU upgrades. I > had an Intel 486/DX266 and upgraded to a AMD 586/133 from Ganberry via > Micro Warehouse. For $140 I have a machine that runs about 2.5 x > faster! > > And no problems with BSD! Yeah! > > I was a little worried having read reports of problems with Cyrix > chips, but so far, so good! If I notice anything, I'll report it. > > Matt > -- Just to add another data point... I've got a 486 box at home that I also upgraded to an AMD 586/133 from an AMD DX/2 at 80 MHz. Since my motherboard could provide 3.3V and it has a not too ancient BIOS, the packaged upgrades from Gainberry and Evergreen really don't offer any added value over just getting just the 586/133 chip. I paid $60+tax from Electrotex here in Houston. The 586/133 is essentially just a 486 with clock tripling/quadrupling and a bigger writeback cache. My no-name VESA motherboard/BIOS doesn't have documented support for writeback nor clock quadrupling. Despite suggestions from helpful hardware hackers that I dike off the appropriate pin on the CPU package (See the datasheet.) to get it to quadruple, I decided to take a more conservative approach. Therefore I'm running the system in writethrough mode and clock tripled mode over a 40 MHz local bus, i.e. the CPU core is running at 120. I am paying a penalty for using writethrough vs. writeback, but I've been insufficiently motivated to try to uncover an undocumented writeback mode. As expected, performance on CPU intensive stuff is therefore a bit over 50% better (Remember the bigger cache than it was before. See the AMD web page for documentation, including the datasheet. The system has been rock solid for about 3 months running FreeBSD or Windows. -- Rob