From owner-freebsd-doc@FreeBSD.ORG Mon Aug 15 19:45:54 2005 Return-Path: X-Original-To: freebsd-doc@freebsd.org Delivered-To: freebsd-doc@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id DF6FC16A41F; Mon, 15 Aug 2005 19:45:54 +0000 (GMT) (envelope-from cracauer@schlepper.zs64.net) Received: from schlepper.zs64.net (schlepper.zs64.net [212.12.50.230]) by mx1.FreeBSD.org (Postfix) with ESMTP id 1385243D45; Mon, 15 Aug 2005 19:45:53 +0000 (GMT) (envelope-from cracauer@schlepper.zs64.net) Received: from schlepper.zs64.net (schlepper [212.12.50.230]) by schlepper.zs64.net (8.13.1/8.12.9) with ESMTP id j7FJjqNU095775; Mon, 15 Aug 2005 21:45:52 +0200 (CEST) (envelope-from cracauer@schlepper.zs64.net) Received: (from cracauer@localhost) by schlepper.zs64.net (8.13.1/8.12.9/Submit) id j7FJjqlP095774; Mon, 15 Aug 2005 15:45:52 -0400 (EDT) (envelope-from cracauer) Date: Mon, 15 Aug 2005 15:45:52 -0400 From: Martin Cracauer To: freebsd-amd64@freebsd.org Message-ID: <20050815154552.A95576@cons.org> Mime-Version: 1.0 Content-Type: multipart/mixed; boundary="PEIAKu/WMn1b1Hv9" Content-Disposition: inline User-Agent: Mutt/1.2.5i Cc: freebsd-doc@freebsd.org Subject: Update for www/amd64/motherboards.sgml - please review X-BeenThere: freebsd-doc@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Documentation project List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 15 Aug 2005 19:45:55 -0000 --PEIAKu/WMn1b1Hv9 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Can somebody who has the SGML system running and is a native speaker of English please review the appended diff? It contains entries for the 4 mainboards that I tried with FreeBSD/AMD64 recently. Thanks Martin --PEIAKu/WMn1b1Hv9 Content-Type: text/plain; charset=us-ascii Content-Disposition: attachment; filename=diff Content-Transfer-Encoding: quoted-printable Index: motherboards.sgml =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D RCS file: /home/dcvs/www/en/platforms/amd64/motherboards.sgml,v retrieving revision 1.36 diff -u -r1.36 motherboards.sgml --- motherboards.sgml 20 Jul 2005 21:39:52 -0000 1.36 +++ motherboards.sgml 15 Aug 2005 19:40:26 -0000 @@ -46,6 +46,42 @@ Integrated NIC does not work, otherwise fully functional. + Abit + + KV8-Pro + VIA K8T800 / Socket 754 + Martin Cracauer + 6-betas, 7-current + On-board ethernet controller (vge(4)) does not work unless stati= cally compiled into kernel. Temperature and fans monitors are connected to = a custom chip, not the Winbond chip they have for voltage. They don't give = out documentation on this custom "uGuru" chip, so no Linux or FreeBSD monit= oring. Sound works just fine. No ECC memory support. I replaced it with an = Asus K8V-X which has working monitoring and ECC support. I would recommend = not giving money to people who needlessly invent cute chips and tell Linux = and FreeBSD to go stuff themself when they want to support these chips. Con= trary to common believe this is also not a very good overclocker. You need = a non-public custom BIOS to even get a working AGP/PCI lock, although the w= hole point of the K8T800-pro chipset is the ability to have that lock. + + + AccelerTech + + ATO2082 + AMD 8111/8131 chipset, dual socket 940 + Martin Cracauer + 6-betas, 7-current + Different name for the Armima Rioworks HDAMB board, see my comme= nts there. + + + Arima + + Rioworks HDAMB + AMD 8111/8131 chipset, dual socket 940 + Martin Cracauer + 6-betas, 7-current on August 2005 + Also sold under the name "AccelerTech ATO2082". Works fine. Soun= d works out of box, as do USB, firewire and SATA. Ethernet only works if co= mpiled statically into the kernel, not as a module. Memory seems slow, I ha= ve PC2100 but it is even slower than it should be with that. Temperature mo= nitoring seems broken. Suspend-to-RAM does not work. This is one of the few= normal-size ATX dual Opteron boards so that you can use a normal case. Can= be had for $120 or so on ebay. Revisions up to C0 or so do not support DDR= 400 memory (PC3200), only DDR333 (PC2700). No dual-core support. Most likel= y no support for E stepping Opterons. Note just in case: like all socket 94= 0 bords it needs registered memory. Nothing fancy or fast but a nice board = if you want it small, don't need PCI-X or PCIe and want full FreeBSD suppor= t. + + + Asus + + KVN-Deluxe + NVidia NForce 3 250 GB / Socket 754 + Martin Cracauer + 6-betas, 7-current + Don't buy this. Ethernet needs the reimplementation of a reverse= engineered Linux driver and didn't do 1000mps for me. Sound would not work= . Under Windows, the AGP port has hangs/stutter with NVidia 6x00 video card= s, in case you plan on dual-purpose with gaming. Lots of issues with Windoz= e drivers, too. Sometimes crashes hard after using PCI TV card, requiring p= ysical (PSU) powercycle. Not very fast for compilation, my Via boards are f= aster (see other entries). However, this board has the Ethernet not on the = PCI bus and under Linux (where the driver works) you can get insane speeds = of ethernet and disk, at the same time. Has ECC support, too. Still, FreeBS= D was no joy on this board. + + Asus A8N-SLI Deluxe @@ -101,6 +137,13 @@ 5.4-STABLE On-board audio untested. The on-board sk0 ethernet controller h= ad issues with 5.3-RELEASE. + Asus + + K8V-X + VIA K8T800 / Socket 754 + Martin Cracauer + 6-betas, 7-current + On-board audio works fine. The on-board sk0 ethernet works but = is slow at about 33 MB/sec. Linux does 70 MB/sec reading, 48 MB writing on = the same card so it's no hardware issue. ECC support. Drives your DDR400 me= mory in DDR333 if you have more than one module and corrupts memory if you = force it into DDR400. Only two fan connectors. I like it better than the A= bit KV8-Pro and that Asus K8N-E deluxe disgrace I had. WARNING: in the defa= ult BIOS shipped switching back and forth between forcing DDR400 and autode= tection, it would not reliably autodetect. If you go from forced memory spe= cs to autodetection load the whole BIOS-wide factory defaults first. This i= s a memory corruption issue so don't ignore this warning. Or don't play wit= h the RAM settings to start from, there was nothing to be gained here. The = board doesn't do DDR400 with two modules and it doesn't have a AGP/PCI lock= so both overclocking and RAM tuning are out. Just plug it in and leave it = alone works best here. Overall, while I list my nitpicks, this is probably = the best socket 754 board for reliable operations and complete FreeBSD supp= ort, with ECC, temperature monitoring etc. Celestica A2210-SCSI --PEIAKu/WMn1b1Hv9--