From <@irisserv.ivw.uni-kl.de:Hentzel@ivw.uni-kl.de> Thu Mar 21 06:23:09 1996 Received: from hairball.ecst.csuchico.edu (root@hairball.ecst.csuchico.edu [132.241.1.108]) by spam.ecst.csuchico.edu (8.7.3/8.7.3) with SMTP id GAA15299 for ; Thu, 21 Mar 1996 06:23:03 -0800 (PST) Received: from uni-kl.de (stepsun.uni-kl.de [131.246.136.50]) by hairball.ecst.csuchico.edu (8.6.12/8.6.12) with SMTP id GAA17519 for ; Thu, 21 Mar 1996 06:22:53 -0800 Received: from irisserv.ivw.uni-kl.de by stepsun.uni-kl.de id aa23636; 21 Mar 96 15:20 MET Received: from localhost by ivw.uni-kl.de with smtp (Smail3.1.28.1 #5) id m0tzlI1-000MncC; Thu, 21 Mar 96 15:24 MET Date: Thu, 21 Mar 96 15:24 MET Newsgroups: alt.comp.periphs.mainboard.asus From: Markus Hentzel To: dranch@ecst.csuchico.edu Subject: Re: Setting for AMD 5x86-133 & GX4 Organization: IVW Message-ID: <4isdkg$duo_001@ivw.uni-kl.de> References: <4ib614$k82@ustsu10.ust.hk> <4ikum3$2l6@charnel.ecst.csuchico.edu> Status: RO X-Status: In article <4ikum3$2l6@charnel.ecst.csuchico.edu>, dranch@ecst.csuchico.edu (David A. Ranch) wrote: MH>Yuen Ming Hon wrote: MH> MH>>Did anyone configurate the AMD 5x86-133 and VL/I 486SV2GX4 successfully? MH>>What is the BIOS and jumpper setting for this combination? MH>>It need not to overclock the CPU, just enough to run at full rated speed. MH> MH>I originally posted this and I still haven't received any responce!! MH>Come on people.. someone must have tried the GX4 with the 133Mhz processor. MH> Hi ! You'll find the jumper settings at Asus' webserver: www.asus.com.tw Here are the settings for 3*40Mhz that have been posted in this group: Configuration: Board: ASUS VL/I-486SV2GX4 Rev 2.0 CPU: AMD Am486-120 (SV8B) CPU BIOS: Award BIOS 401A0-0401 L2 Cache: 32k*8 * 8 (15nsec) DRAM: One 16 Mb SIMM (60 nsec) Jumper settings: clock frequency set to 40Mhz 1 VL-Bus wait state JUMPER | ----------------+-------------------- CPU: | JP 5 | 1&2 JP 6 | 1&2 both short CPU selection: | JP 11 | 2&3 short | Short for: | JP 16: | 1&2 5&6 JP 17: | 1&2 5&6 JP 18: | 1&2 JP 19: | 1&2 3&4 5&6 JP 20: | open JP 21: | 3&4 for L1 WB (DMA-problem --> NO floppy access) | 1&2 for L1 WT (ok) JP 22: | 1&2 4&5 BIOS settings: As fast as possible (NO waits, in both cases) It does'nt make any difference how i set the L2 Cache. (OFF/L2-WT/L2-WB[with or without dirty]) Hope it'll help you. Ciao Markus --- hentzel@ivw.uni-kl.de From roman@Galaxy.Net Fri Mar 22 16:25:14 1996 Received: from mail.galaxy.net (root@mars.Galaxy.net [198.138.93.17]) by spam.ecst.csuchico.edu (8.7.3/8.7.3) with SMTP id QAA17834 for ; Fri, 22 Mar 1996 16:25:12 -0800 (PST) Received: from zaphod by mail.galaxy.net (SMI-8.6/SMI-SVR4) id TAA11167; Fri, 22 Mar 1996 19:25:06 -0500 Date: Fri, 22 Mar 1996 19:25:01 -0500 (EST) From: Roman Shapiro X-Sender: roman@zaphod To: "David A. Ranch" Subject: Re: Setting for AMD 5x86-133 & GX4 In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Status: RO X-Status: > Can I get you to let me know what your settings are for the: > > - motherboard jumper settings I'm not at home, so don't have access to the machine or the manual. If you have the manual, I can tell you which entry to use (like AMD Enhanced). Let me know if you do, and I'll mail you the info. I really don't have the time to draw a diagram or something :). > - BIOS wait states / speed settings With the chip running at 4x40, I have the speed to fastest and fewest wait states (I'm pretty sure). But this is something that you have to tweak in the end by yourself. At 4x33, you should be able to set things to max speed. > > > Also.. did you upgrade your BIOS yourself (v401?) or did you buy it from > ASUS direct? > The guy who sold me the board (Dave Gee) sent me a new bios chip (401) in exchange for the old bios chip (no charge). From roman@Galaxy.Net Thu Mar 28 11:27:27 1996 Received: from mail.galaxy.net (root@mars.Galaxy.net [198.138.93.17]) by spam.ecst.csuchico.edu (8.7.3/8.7.3) with SMTP id LAA14379 for ; Thu, 28 Mar 1996 11:27:14 -0800 (PST) Received: from zaphod by mail.galaxy.net (SMI-8.6/SMI-SVR4) id OAA00691; Thu, 28 Mar 1996 14:27:06 -0500 Date: Thu, 28 Mar 1996 14:27:01 -0500 (EST) From: Roman Shapiro X-Sender: roman@zaphod To: "David A. Ranch" Subject: Re: Setting for AMD 5x86-133 & GX4 : I've got the processor now! In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Status: RO X-Status: Here's something I read in a post recently: ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Hi I've seen numerous requests for information on how to set the jumpers correctly for this combination. I received two of these processors this weekend, and managed to get them going using the following jumper settings: JP16: 1-2 4-5 JP17: 1-2 5-6 JP18: 1-2 JP19: 1-2 3-4 5-6 JP20: 1-2 JP21: 2-3 JP22: 1-2 4-5 If you leave JP20 open the CPU runs in clock trebling mode instead of clock quadrupling mode. You could experiment and try to get it to work as a 3*50MHz system. Both CPUs ran flawlessly at both 133MHz (4*33MHz) and 160MHz (4*40MHz). The only problem I did encounter was that QEMM 7.5 hung when it started, no matter what the speed of the CPU was. I think it has something to do with QEMM identifying the CPU incorrectly. Cheers Graham Inggs +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ On Thu, 28 Mar 1996, David A. Ranch wrote: > > Hey Roman, > > Welp, I broke down and bought a 133Mhz processor. I'm in the process of > burning in the v401 BIOS and I would to pick your brain about what the > jumper settings are for the GX4 motherboard. Again, thanks for the help > and I plan on posting all my results to the ASUS newsgroup. > > (from you earilier) > > > If you have the manual, I can tell you which entry to use (like AMD > > Enhanced). > > Ok, the first thing is.. this processor is called the 5x86-133. The Asus > WWW site keeps mentioning two AMD DX4 processors: SV8B and the NV8T. > I'm under the impression that this 133Mhz processor is neither. Can you > clear this up for me? Ok.. for the processor stuff, check out the bottom > of this e-mail since I'll put all the differentGx4 processor types there > and I'll include the readme file that came with the bios upgrade just in > case. > The SV8 and SV8 stuff only applied to the 80-120 Mhz processors. The 133 has and ADZ somewhere I think. > > With the chip running at 4x40, I have the speed to fastest and fewest > > wait states (I'm pretty sure). But this is something that you have to > > tweak in the end by yourself. At 4x33, you should be able to set things > > to max speed. > > Great.. I'm running an AMD Dx/2 80 5V now and I can only run it at FASTER. > I still might not be able to get my system to run at fastest but bah.. > thats ok. Two question(s): > > 1) ASUS says that you MUST run the L1 cache in > Write-Through mode instead of the preferable Write-Back mode if > you have a bus-mastering card like a SCSI card (which I do). > Have you had to deal with this or it doesn't apply for your > situation? > Not sure. I use write-back and have everything except for a SCSI card. > 2) What settings are you using for your VLBus? > > - VESA delay ENABLED? (ASUS default is enabled) > - <33Mhz & 0ws (This is what I'm running now) > or > >33Mhz & 1ws > At 40Mhz, I have the VESA delay enabled. It may work without it at 33Mhz. > > OK.. lastly,,from the manual as you requested, here are the different CPU > jumper settings. Could you note to me which I need to run the processor > at both 133Mhz and 160Mhz for our GX4 motherboards? > The same setting should allow you to run at 133 and 160- you just have to change the bus speed (from 33 to 40). Compare the settings from the beginning of this message to the differenct cpu setting from the manual. I think it's closest to the AMD486DXL4 - 2x clock- you may not have to change the jumpers much. :) There may be one jumper different- the one that enables internal write-back. Also, be warned that at 160Mhz (4x40), my GX4 boots up and says 150Mhz. Good luck. -Roman > --- > > probably no - SL Enhanced 486SX & SX2 or Non-SL 486 & SX2 > > probably no - SL Enhanced (non-SL) 486DX, DX2, 487SX & ODP; 486Dx/4 @ > 3x clock > > probably no - 486DX4 - 2.5x clock > > probably no - 486DX4 - 2x clock (BIOS reports AM486DXL4 3x) > > possibly - P24D Write-Back L1 cache > JP21 Short 2&3 - Write-Through L1 cache > > possibly - P24CT - Write-back L1 cache > JP21 Short 2&3 - Write-Through L1 cache > > possibly - P24T - Write-back L1 cache > JP21 Short 2&3 - Write-Through L1 cache > 237-pin SL ODP > > possibly 120Mhz - 486DX4 ODP (OverDrive Processor?) - 3x clock > > possibly - 486DX4 ODP (OverDrive Processor?) - 2.5 clock > > doubtful - 486DX4 ODP (OverDrive Processor?) - 2.0 clock > > doubtful - AMD486SXL, AMD486SX2L > > doubtful - AMD486DXl, AMD486DX2L > > possible 120Mhz - AMD486DXL4 - 3x clock > > This is what I - AMD486DXL4 - 2x clock > run my Dx/2 80 > at > > doubtful - Cyrix 486DX, DX2 > > doubtful - Cyrix 486DX2-V > > doubtful - Cyrix DX5 - 3x clock > > doubtful - Cyrix DX5 - 2x clock > > > ======= > > > This is the text file that came with the BIOS upgrade image. I'm > including this just in case it helps. > > ---- > > BIOS file : SV2G0401.AWD > BIOS Ver. : 0401 (#401A0-0401) > Related Mainboard : VL/I-486SV2G(X4) Rev. 2.0 or above > Related Chipset : SiS 85C471 chipset > > > Description : > > 1. To support the following CPUs : > a.) AMD 486DX4 (SV8B) : with green & L1 cache write-back. > b.) Intel 486DX4 (&EW) : with green & L1 cache write-back. > c.) Cyrix 486DX4 : with green & L1 cache write-back. > d.) Cyrix 5x86 (M1sc) : with Pentium code inside. > e.) AMD 486DX4 (NV8T) : non-green & L1 cache write-through. > > The new CPUs jumper setting as the follow : > > A) Cyrix 486DX4 P/O > Cyrix 5x86 (M1sc) > > JP16 : 1-2,5-6 (short) > JP17 : 1-2,5-6 (short) > JP18 : 1-2 (short) > JP19 : 1-2,3-4,5-6 (short) > JP20 : (none) > JP21 : (none) > JP22 : 1-2 (short) > > > B) AMD 486DX4 (SV8B) > Intel 486DX4 (&EW) > > JP16 : 1-2,4-5 (short) > JP17 : 1-2,5-6 (short) > JP18 : 1-2 (short) > JP19 : 1-2,3-4,5-6 (short) > JP20 : (none) > JP21 : 3-4 (short) > JP22 : 1-2,4-5 (short) > > C) Cyrix 486DX4 > > JP16 : 2-3,5-6 (short) > JP17 : 2-3,5-6 (short) > JP18 : 2-3 (short) > JP19 : 2-3 (short) > JP20 : 2-3 (short) > JP21 : 1-2 (short) > JP22 : 1-2 (short) > > D) AMD 486DX4 (NV8T) > AMD 486DXL4 > > JP16 : 1-2,5-6 (short) > JP17 : 4-5 (short) > JP18 : 4-5 (short) > JP19 : (none) > JP20 : 4-5 (short) > JP21 : (2-3 short : 2x CLK, 2-3 open : 3x CLK) > JP22 : 1-2, 4-5 (short) > > > --David > +----------------------------------------------------------------------------+ > | David A. Ranch - ISDN/Cablemodems/PCs dranch@ecst.csuchico.edu | > +---- ----+ > +----- For more detailed info, check out: www.ecst.csuchico.edu/~dranch -----+ > > > From dranch@ecst.csuchico.edu Thu Mar 28 12:21:56 1996 Received: from corpse.ecst.csuchico.edu (root@corpse.ecst.csuchico.edu [132.241.5.10]) by spam.ecst.csuchico.edu (8.7.3/8.7.3) with SMTP id MAA19298 for ; Thu, 28 Mar 1996 12:21:56 -0800 (PST) Received: from localhost (dranch@localhost [127.0.0.1]) by corpse.ecst.csuchico.edu (8.6.12/8.6.12) with SMTP id MAA20481; Thu, 28 Mar 1996 12:21:48 -0800 Date: Thu, 28 Mar 1996 12:21:48 -0800 (PST) From: "David A. Ranch" Reply-To: "David A. Ranch" To: ginggs@iafrica.com cc: "David A. Ranch" Subject: Re: Setting for AMD 5x86-133 & GX4 : I've got the processor now! In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Status: RO X-Status: Hello Gramm, Roman Shapiro (roman@galaxy.net) forwarded me your news post on the ASUS newsgroup about the AMD 133Mhz processor and the GX4 motherboard. I wanted to CC: this to you just in case you had any additional ideas. Yes.. this message is a long one but I plan to post a comprehensive message on the newsgroup once I know -exactly- whats going on.) Anyway, could you let me know your thoughts about the following? (me) > > 1) ASUS says that you MUST run the L1 cache in > > Write-Through mode instead of the preferable Write-Back mode if > > you have a bus-mastering card like a SCSI card (which I do). > > I currently have an Adaptec 2842 Bus-mastering VLB SCSI > > controller and I bet it will have possible problems. > > > > Have you had to deal with this or it doesn't apply for your > > situation? > > > (roman) > Not sure. I use write-back and have everything except for a SCSI card. > > 2) What settings are you using for your VLBus? > > > > - VESA delay ENABLED? (ASUS default is enabled) > > - <33Mhz & 0ws (This is what I'm running now) > > or > > >33Mhz & 1ws > > > At 40Mhz, I have the VESA delay enabled. It may work without it at 33Mhz. > > OK.. lastly,,from the manual as you requested, here are the different CPU > > jumper settings. Could you note to me which I need to run the processor > > at both 133Mhz and 160Mhz for our GX4 motherboards? > > > >(roman) > The same setting (from Gramm's news post) should allow you to run at 133 > and 160- you just have to change the bus speed (from 33 to 40). Compare > the settings from the beginning of this message to the differenct cpu > setting from the manual. I think it's closest to the AMD486DXL4 - 2x > clock- you may not have to change the jumpers much. :) There may be one > jumper different- the one that enables internal write-back. > > Also, be warned that at 160Mhz (4x40), my GX4 boots up and says 150Mhz. > > Good luck. > -Roman Also Gramm, did you upgrade your BIOS yourself? I have the 401 BIOS image here and I'm going to burn it today but I'm not sure what format the image is in. Is it in HEX, Object, etc? Anyway... thank you for your time and I hope to hear from you soon. (The following is the settings in the GX4 motherboard manual and then the readme file included with the BIOS image from ASUS's WWW site.) > > > --- > > > > probably no - SL Enhanced 486SX & SX2 or Non-SL 486 & SX2 > > > > probably no - SL Enhanced (non-SL) 486DX, DX2, 487SX & ODP; 486Dx/4 @ > > 3x clock > > > > probably no - 486DX4 - 2.5x clock > > > > probably no - 486DX4 - 2x clock (BIOS reports AM486DXL4 3x) > > > > possibly - P24D Write-Back L1 cache > > JP21 Short 2&3 - Write-Through L1 cache > > > > possibly - P24CT - Write-back L1 cache > > JP21 Short 2&3 - Write-Through L1 cache > > > > possibly - P24T - Write-back L1 cache > > JP21 Short 2&3 - Write-Through L1 cache > > 237-pin SL ODP > > > > possibly 120Mhz - 486DX4 ODP (OverDrive Processor?) - 3x clock > > > > possibly - 486DX4 ODP (OverDrive Processor?) - 2.5 clock > > > > doubtful - 486DX4 ODP (OverDrive Processor?) - 2.0 clock > > > > doubtful - AMD486SXL, AMD486SX2L > > > > doubtful - AMD486DXl, AMD486DX2L > > > > possible 120Mhz - AMD486DXL4 - 3x clock > > > > This is what I - AMD486DXL4 - 2x clock > > run my Dx/2 80 > > at > > > > doubtful - Cyrix 486DX, DX2 > > > > doubtful - Cyrix 486DX2-V > > > > doubtful - Cyrix DX5 - 3x clock > > > > doubtful - Cyrix DX5 - 2x clock > > > > > > ======= > > > > > > This is the text file that came with the BIOS upgrade image. I'm > > including this just in case it helps. > > > > ---- > > > > BIOS file : SV2G0401.AWD > > BIOS Ver. : 0401 (#401A0-0401) > > Related Mainboard : VL/I-486SV2G(X4) Rev. 2.0 or above > > Related Chipset : SiS 85C471 chipset > > > > > > Description : > > > > 1. To support the following CPUs : > > a.) AMD 486DX4 (SV8B) : with green & L1 cache write-back. > > b.) Intel 486DX4 (&EW) : with green & L1 cache write-back. > > c.) Cyrix 486DX4 : with green & L1 cache write-back. > > d.) Cyrix 5x86 (M1sc) : with Pentium code inside. > > e.) AMD 486DX4 (NV8T) : non-green & L1 cache write-through. > > > > The new CPUs jumper setting as the follow : > > > > A) Cyrix 486DX4 P/O > > Cyrix 5x86 (M1sc) > > > > JP16 : 1-2,5-6 (short) > > JP17 : 1-2,5-6 (short) > > JP18 : 1-2 (short) > > JP19 : 1-2,3-4,5-6 (short) > > JP20 : (none) > > JP21 : (none) > > JP22 : 1-2 (short) > > > > > > B) AMD 486DX4 (SV8B) > > Intel 486DX4 (&EW) > > > > JP16 : 1-2,4-5 (short) > > JP17 : 1-2,5-6 (short) > > JP18 : 1-2 (short) > > JP19 : 1-2,3-4,5-6 (short) > > JP20 : (none) > > JP21 : 3-4 (short) > > JP22 : 1-2,4-5 (short) > > > > C) Cyrix 486DX4 > > > > JP16 : 2-3,5-6 (short) > > JP17 : 2-3,5-6 (short) > > JP18 : 2-3 (short) > > JP19 : 2-3 (short) > > JP20 : 2-3 (short) > > JP21 : 1-2 (short) > > JP22 : 1-2 (short) > > > > D) AMD 486DX4 (NV8T) > > AMD 486DXL4 > > > > JP16 : 1-2,5-6 (short) > > JP17 : 4-5 (short) > > JP18 : 4-5 (short) > > JP19 : (none) > > JP20 : 4-5 (short) > > JP21 : (2-3 short : 2x CLK, 2-3 open : 3x CLK) > > JP22 : 1-2, 4-5 (short) > > --David +----------------------------------------------------------------------------+ | David A. Ranch - ISDN/Cablemodems/PCs dranch@ecst.csuchico.edu | +---- ----+ +----- For more detailed info, check out: www.ecst.csuchico.edu/~dranch -----+ From dranch@ecst.csuchico.edu Thu Mar 28 12:27:01 1996 Received: from corpse.ecst.csuchico.edu (root@corpse.ecst.csuchico.edu [132.241.5.10]) by spam.ecst.csuchico.edu (8.7.3/8.7.3) with SMTP id MAA19672 for ; Thu, 28 Mar 1996 12:26:59 -0800 (PST) Received: from localhost (dranch@localhost [127.0.0.1]) by corpse.ecst.csuchico.edu (8.6.12/8.6.12) with SMTP id MAA20978; Thu, 28 Mar 1996 12:26:52 -0800 Date: Thu, 28 Mar 1996 12:26:51 -0800 (PST) From: "David A. Ranch" To: rlindsay@jumpnet.com cc: "David A. Ranch" Subject: Re: ASUS GX4 BIOS Upgrade Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Status: RO X-Status: Hey Rick, I read a post from you on the ASUS newsgroup about upgrading the BIOS on a ASUS GX4 motherboard. I have the Award 401 image and have a EPROM burner ready. The problem is that the included text file doesn't tell the user which programmer file format to use! I'm going to try HEX first but if you have any ideas or comments, I would greatly appreciate it! --David +----------------------------------------------------------------------------+ | David A. Ranch - ISDN/Cablemodems/PCs dranch@ecst.csuchico.edu | +---- ----+ +----- For more detailed info, check out: www.ecst.csuchico.edu/~dranch -----+ From dranch@ecst.csuchico.edu Thu Mar 28 12:31:47 1996 Received: from corpse.ecst.csuchico.edu (root@corpse.ecst.csuchico.edu [132.241.5.10]) by spam.ecst.csuchico.edu (8.7.3/8.7.3) with SMTP id MAA20129 for ; Thu, 28 Mar 1996 12:31:46 -0800 (PST) Received: from localhost (dranch@localhost [127.0.0.1]) by corpse.ecst.csuchico.edu (8.6.12/8.6.12) with SMTP id MAA21133; Thu, 28 Mar 1996 12:30:05 -0800 Date: Thu, 28 Mar 1996 12:30:04 -0800 (PST) From: "David A. Ranch" To: difilipp@whidbey.net cc: "David A. Ranch" Subject: Re: Setting for AMD 5x86-133 & GX4 : I've got the processor now! (fwd) Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Status: RO X-Status: Hey John, I'm in the process of upgrading my BIOS and getting the 133Mhz processor to run at 160Mhz. Here is my current thread to two other people about the process. Once I have ALL the details, I plan on making a master GX4 upgrade document and posting it on the ASUS newsgroup. Could you give the technical specs a once-over and let me know if you dis-agree with anything? I greatly appreciate it and I'll talk to you soon! --David +----------------------------------------------------------------------------+ | David A. Ranch - ISDN/Cablemodems/PCs dranch@ecst.csuchico.edu | +---- ----+ +----- For more detailed info, check out: www.ecst.csuchico.edu/~dranch -----+ ---------- Forwarded message ---------- Date: Thu, 28 Mar 1996 14:27:01 -0500 (EST) From: Roman Shapiro To: "David A. Ranch" Subject: Re: Setting for AMD 5x86-133 & GX4 : I've got the processor now! Here's something I read in a post recently: ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Hi I've seen numerous requests for information on how to set the jumpers correctly for this combination. I received two of these processors this weekend, and managed to get them going using the following jumper settings: JP16: 1-2 4-5 JP17: 1-2 5-6 JP18: 1-2 JP19: 1-2 3-4 5-6 JP20: 1-2 JP21: 2-3 JP22: 1-2 4-5 If you leave JP20 open the CPU runs in clock trebling mode instead of clock quadrupling mode. You could experiment and try to get it to work as a 3*50MHz system. Both CPUs ran flawlessly at both 133MHz (4*33MHz) and 160MHz (4*40MHz). The only problem I did encounter was that QEMM 7.5 hung when it started, no matter what the speed of the CPU was. I think it has something to do with QEMM identifying the CPU incorrectly. Cheers Graham Inggs +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ On Thu, 28 Mar 1996, David A. Ranch wrote: > > Hey Roman, > > Welp, I broke down and bought a 133Mhz processor. I'm in the process of > burning in the v401 BIOS and I would to pick your brain about what the > jumper settings are for the GX4 motherboard. Again, thanks for the help > and I plan on posting all my results to the ASUS newsgroup. > > (from you earilier) > > > If you have the manual, I can tell you which entry to use (like AMD > > Enhanced). > > Ok, the first thing is.. this processor is called the 5x86-133. The Asus > WWW site keeps mentioning two AMD DX4 processors: SV8B and the NV8T. > I'm under the impression that this 133Mhz processor is neither. Can you > clear this up for me? Ok.. for the processor stuff, check out the bottom > of this e-mail since I'll put all the differentGx4 processor types there > and I'll include the readme file that came with the bios upgrade just in > case. > The SV8 and SV8 stuff only applied to the 80-120 Mhz processors. The 133 has and ADZ somewhere I think. > > With the chip running at 4x40, I have the speed to fastest and fewest > > wait states (I'm pretty sure). But this is something that you have to > > tweak in the end by yourself. At 4x33, you should be able to set things > > to max speed. > > Great.. I'm running an AMD Dx/2 80 5V now and I can only run it at FASTER. > I still might not be able to get my system to run at fastest but bah.. > thats ok. Two question(s): > > 1) ASUS says that you MUST run the L1 cache in > Write-Through mode instead of the preferable Write-Back mode if > you have a bus-mastering card like a SCSI card (which I do). > Have you had to deal with this or it doesn't apply for your > situation? > Not sure. I use write-back and have everything except for a SCSI card. > 2) What settings are you using for your VLBus? > > - VESA delay ENABLED? (ASUS default is enabled) > - <33Mhz & 0ws (This is what I'm running now) > or > >33Mhz & 1ws > At 40Mhz, I have the VESA delay enabled. It may work without it at 33Mhz. > > OK.. lastly,,from the manual as you requested, here are the different CPU > jumper settings. Could you note to me which I need to run the processor > at both 133Mhz and 160Mhz for our GX4 motherboards? > The same setting should allow you to run at 133 and 160- you just have to change the bus speed (from 33 to 40). Compare the settings from the beginning of this message to the differenct cpu setting from the manual. I think it's closest to the AMD486DXL4 - 2x clock- you may not have to change the jumpers much. :) There may be one jumper different- the one that enables internal write-back. Also, be warned that at 160Mhz (4x40), my GX4 boots up and says 150Mhz. Good luck. -Roman > --- > > probably no - SL Enhanced 486SX & SX2 or Non-SL 486 & SX2 > > probably no - SL Enhanced (non-SL) 486DX, DX2, 487SX & ODP; 486Dx/4 @ > 3x clock > > probably no - 486DX4 - 2.5x clock > > probably no - 486DX4 - 2x clock (BIOS reports AM486DXL4 3x) > > possibly - P24D Write-Back L1 cache > JP21 Short 2&3 - Write-Through L1 cache > > possibly - P24CT - Write-back L1 cache > JP21 Short 2&3 - Write-Through L1 cache > > possibly - P24T - Write-back L1 cache > JP21 Short 2&3 - Write-Through L1 cache > 237-pin SL ODP > > possibly 120Mhz - 486DX4 ODP (OverDrive Processor?) - 3x clock > > possibly - 486DX4 ODP (OverDrive Processor?) - 2.5 clock > > doubtful - 486DX4 ODP (OverDrive Processor?) - 2.0 clock > > doubtful - AMD486SXL, AMD486SX2L > > doubtful - AMD486DXl, AMD486DX2L > > possible 120Mhz - AMD486DXL4 - 3x clock > > This is what I - AMD486DXL4 - 2x clock > run my Dx/2 80 > at > > doubtful - Cyrix 486DX, DX2 > > doubtful - Cyrix 486DX2-V > > doubtful - Cyrix DX5 - 3x clock > > doubtful - Cyrix DX5 - 2x clock > > > ======= > > > This is the text file that came with the BIOS upgrade image. I'm > including this just in case it helps. > > ---- > > BIOS file : SV2G0401.AWD > BIOS Ver. : 0401 (#401A0-0401) > Related Mainboard : VL/I-486SV2G(X4) Rev. 2.0 or above > Related Chipset : SiS 85C471 chipset > > > Description : > > 1. To support the following CPUs : > a.) AMD 486DX4 (SV8B) : with green & L1 cache write-back. > b.) Intel 486DX4 (&EW) : with green & L1 cache write-back. > c.) Cyrix 486DX4 : with green & L1 cache write-back. > d.) Cyrix 5x86 (M1sc) : with Pentium code inside. > e.) AMD 486DX4 (NV8T) : non-green & L1 cache write-through. > > The new CPUs jumper setting as the follow : > > A) Cyrix 486DX4 P/O > Cyrix 5x86 (M1sc) > > JP16 : 1-2,5-6 (short) > JP17 : 1-2,5-6 (short) > JP18 : 1-2 (short) > JP19 : 1-2,3-4,5-6 (short) > JP20 : (none) > JP21 : (none) > JP22 : 1-2 (short) > > > B) AMD 486DX4 (SV8B) > Intel 486DX4 (&EW) > > JP16 : 1-2,4-5 (short) > JP17 : 1-2,5-6 (short) > JP18 : 1-2 (short) > JP19 : 1-2,3-4,5-6 (short) > JP20 : (none) > JP21 : 3-4 (short) > JP22 : 1-2,4-5 (short) > > C) Cyrix 486DX4 > > JP16 : 2-3,5-6 (short) > JP17 : 2-3,5-6 (short) > JP18 : 2-3 (short) > JP19 : 2-3 (short) > JP20 : 2-3 (short) > JP21 : 1-2 (short) > JP22 : 1-2 (short) > > D) AMD 486DX4 (NV8T) > AMD 486DXL4 > > JP16 : 1-2,5-6 (short) > JP17 : 4-5 (short) > JP18 : 4-5 (short) > JP19 : (none) > JP20 : 4-5 (short) > JP21 : (2-3 short : 2x CLK, 2-3 open : 3x CLK) > JP22 : 1-2, 4-5 (short) > > > --David > +----------------------------------------------------------------------------+ > | David A. Ranch - ISDN/Cablemodems/PCs dranch@ecst.csuchico.edu | > +---- ----+ > +----- For more detailed info, check out: www.ecst.csuchico.edu/~dranch -----+ > > > From difilipp@whidbey.net Thu Mar 28 17:18:56 1996 Received: from islander.whidbey.net (root@islander.whidbey.net [204.57.140.1]) by spam.ecst.csuchico.edu (8.7.3/8.7.3) with ESMTP id RAA14948 for ; Thu, 28 Mar 1996 17:18:51 -0800 (PST) Received: from difilippo (asn63.whidbey.net [204.57.140.133]) by islander.whidbey.net (8.7.5/8.7.3) with SMTP id RAA09669 for ; Thu, 28 Mar 1996 17:14:27 -0800 (PST) Message-Id: <2.2.32.19960329011829.00699eac@whidbey.net> X-Sender: difilipp@whidbey.net X-Mailer: Windows Eudora Pro Version 2.2 (32) Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Thu, 28 Mar 1996 17:18:29 -0800 To: "David A. Ranch" From: John and Nellena Difilippo Subject: Re: Setting for AMD 5x86-133 & GX4 : I've got the processor now! (fwd) Status: RO X-Status: A GX4 rev 2.0 Award 4.50G version 0306 Promise EIDE 2300 Plus controller Diamond Stealth 64 Video series 2000 (2 Meg Dram) 16 Meg Ram (70 nsec) The settings I came up with are based on the AMD Enhanced 120's. The cpu's pin assignments are basically the same, the only difference seems to be what the the state of the input does to the chip. (clock multiplication and cache settings). I'm confident I've enabled L1 write back cache. These settings are still preliminary as I'm still waiting for the latest bios, but at present I'm running overclocked at 160Mhz and this $70 dollar chip is outperforming the $270 Pentium Overdrive 83 Mhz chip in every category including CPU 16 and 32 mark. (using winbench 96). I'm not sure it will last at 160Mhz, but the replacement of it won't be that expensive if it does give in. I know the next one will always work at 133Mhz. If these settings can be changed, I'm welcome to suggestions as I don't claim to have all of the answers. They are proven on my system. Bios settings: Auto Config Disabled Fast Reset Emul Enabled AT Bus Clock 1/4 CLKIN(1/5 for 160 Mhz) DRAM Speed Fastest DRAM Write WS 0WS DRAM Write CAS 1T DRAM Write Burst Enabled Slow Refresh Disabled Hidden Refresh Enabled L2 Cache Scheme Write Back L1 Cache Scheme Write Back Cache Burst Read 1T Cache Write Cycle 2T Video Shadow Non-cache Memory Hole Disabled Fast Reset Latency 2 usec Latch Local Bus T2 Local Bus Ready Transparent for 133Mhz Sychronize for 160Mhz Jumper Settings: JP5 1+2 short JP6 1+2 short JP11 2+3 short JP16 1+2, 4+5 short JP17 1+2, 5+6 short JP18 1+2 short JP19 2+3, 4+5 short JP20 1+2 short for 4x clock, All open for 3X clock JP21 3+4 short (should be write back cache), Open for write thru JP22 1+2, 4+5 short JP23, JP24, JP25 33Mhz for 133 operation, 40Mhz for 160 operation JP26 2+3 short JP28 1+2 short for 133Mhz, 2+3 short for 160Mhz JP29 1+2 short JP32 1+2 short Just let me know if this worked for you. ASUS can give me a job! difilipp@whidbey.net At 12:30 PM 3/28/96 -0800, you wrote: > >Hey John, > >I'm in the process of upgrading my BIOS and getting the 133Mhz processor >to run at 160Mhz. Here is my current thread to two other people about the >process. Once I have ALL the details, I plan on making a master GX4 >upgrade document and posting it on the ASUS newsgroup. > >Could you give the technical specs a once-over and let me know if you >dis-agree with anything? > >I greatly appreciate it and I'll talk to you soon! > > >--David > +----------------------------------------------------------------------------+ > | David A. Ranch - ISDN/Cablemodems/PCs dranch@ecst.csuchico.edu | > +---- ----+ > +----- For more detailed info, check out: www.ecst.csuchico.edu/~dranch -----+ > >---------- Forwarded message ---------- >Date: Thu, 28 Mar 1996 14:27:01 -0500 (EST) >From: Roman Shapiro >To: "David A. Ranch" >Subject: Re: Setting for AMD 5x86-133 & GX4 : I've got the processor now! > >Here's something I read in a post recently: > >++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ >Hi >I've seen numerous requests for information on how to set the jumpers >correctly for this combination. I received two of these processors this >weekend, and managed to get them going using the following jumper >settings: > >JP16: 1-2 4-5 >JP17: 1-2 5-6 >JP18: 1-2 >JP19: 1-2 3-4 5-6 >JP20: 1-2 >JP21: 2-3 >JP22: 1-2 4-5 > >If you leave JP20 open the CPU runs in clock trebling mode instead of >clock quadrupling mode. You could experiment and try to get it to work >as a 3*50MHz system. Both CPUs ran flawlessly at both 133MHz (4*33MHz) >and 160MHz (4*40MHz). The only problem I did encounter was that QEMM >7.5 hung when it started, no matter what the speed of the CPU was. >I think it has something to do with QEMM identifying the CPU incorrectly. >Cheers >Graham Inggs >+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ > > > >On Thu, 28 Mar 1996, David A. Ranch wrote: > >> >> Hey Roman, >> >> Welp, I broke down and bought a 133Mhz processor. I'm in the process of >> burning in the v401 BIOS and I would to pick your brain about what the >> jumper settings are for the GX4 motherboard. Again, thanks for the help >> and I plan on posting all my results to the ASUS newsgroup. >> >> (from you earilier) >> >> > If you have the manual, I can tell you which entry to use (like AMD >> > Enhanced). >> >> Ok, the first thing is.. this processor is called the 5x86-133. The Asus >> WWW site keeps mentioning two AMD DX4 processors: SV8B and the NV8T. >> I'm under the impression that this 133Mhz processor is neither. Can you >> clear this up for me? Ok.. for the processor stuff, check out the bottom >> of this e-mail since I'll put all the differentGx4 processor types there >> and I'll include the readme file that came with the bios upgrade just in >> case. >> >The SV8 and SV8 stuff only applied to the 80-120 Mhz processors. The 133 >has and ADZ somewhere I think. > >> > With the chip running at 4x40, I have the speed to fastest and fewest >> > wait states (I'm pretty sure). But this is something that you have to >> > tweak in the end by yourself. At 4x33, you should be able to set things >> > to max speed. >> >> Great.. I'm running an AMD Dx/2 80 5V now and I can only run it at FASTER. >> I still might not be able to get my system to run at fastest but bah.. >> thats ok. Two question(s): >> >> 1) ASUS says that you MUST run the L1 cache in >> Write-Through mode instead of the preferable Write-Back mode if >> you have a bus-mastering card like a SCSI card (which I do). >> Have you had to deal with this or it doesn't apply for your >> situation? >> >Not sure. I use write-back and have everything except for a SCSI card. > >> 2) What settings are you using for your VLBus? >> >> - VESA delay ENABLED? (ASUS default is enabled) >> - <33Mhz & 0ws (This is what I'm running now) >> or >> >33Mhz & 1ws >> >At 40Mhz, I have the VESA delay enabled. It may work without it at 33Mhz. > >> >> OK.. lastly,,from the manual as you requested, here are the different CPU >> jumper settings. Could you note to me which I need to run the processor >> at both 133Mhz and 160Mhz for our GX4 motherboards? >> > >The same setting should allow you to run at 133 and 160- you just have to >change the bus speed (from 33 to 40). Compare the settings from the >beginning of this message to the differenct cpu setting from the manual. >I think it's closest to the AMD486DXL4 - 2x clock- you may not have to >change the jumpers much. :) There may be one jumper different- the one >that enables internal write-back. > >Also, be warned that at 160Mhz (4x40), my GX4 boots up and says 150Mhz. > >Good luck. >-Roman > >> --- >> >> probably no - SL Enhanced 486SX & SX2 or Non-SL 486 & SX2 >> >> probably no - SL Enhanced (non-SL) 486DX, DX2, 487SX & ODP; 486Dx/4 @ >> 3x clock >> >> probably no - 486DX4 - 2.5x clock >> >> probably no - 486DX4 - 2x clock (BIOS reports AM486DXL4 3x) >> >> possibly - P24D Write-Back L1 cache >> JP21 Short 2&3 - Write-Through L1 cache >> >> possibly - P24CT - Write-back L1 cache >> JP21 Short 2&3 - Write-Through L1 cache >> >> possibly - P24T - Write-back L1 cache >> JP21 Short 2&3 - Write-Through L1 cache >> 237-pin SL ODP >> >> possibly 120Mhz - 486DX4 ODP (OverDrive Processor?) - 3x clock >> >> possibly - 486DX4 ODP (OverDrive Processor?) - 2.5 clock >> >> doubtful - 486DX4 ODP (OverDrive Processor?) - 2.0 clock >> >> doubtful - AMD486SXL, AMD486SX2L >> >> doubtful - AMD486DXl, AMD486DX2L >> >> possible 120Mhz - AMD486DXL4 - 3x clock >> >> This is what I - AMD486DXL4 - 2x clock >> run my Dx/2 80 >> at >> >> doubtful - Cyrix 486DX, DX2 >> >> doubtful - Cyrix 486DX2-V >> >> doubtful - Cyrix DX5 - 3x clock >> >> doubtful - Cyrix DX5 - 2x clock >> >> >> ======= >> >> >> This is the text file that came with the BIOS upgrade image. I'm >> including this just in case it helps. >> >> ---- >> >> BIOS file : SV2G0401.AWD >> BIOS Ver. : 0401 (#401A0-0401) >> Related Mainboard : VL/I-486SV2G(X4) Rev. 2.0 or above >> Related Chipset : SiS 85C471 chipset >> >> >> Description : >> >> 1. To support the following CPUs : >> a.) AMD 486DX4 (SV8B) : with green & L1 cache write-back. >> b.) Intel 486DX4 (&EW) : with green & L1 cache write-back. >> c.) Cyrix 486DX4 : with green & L1 cache write-back. >> d.) Cyrix 5x86 (M1sc) : with Pentium code inside. >> e.) AMD 486DX4 (NV8T) : non-green & L1 cache write-through. >> >> The new CPUs jumper setting as the follow : >> >> A) Cyrix 486DX4 P/O >> Cyrix 5x86 (M1sc) >> >> JP16 : 1-2,5-6 (short) >> JP17 : 1-2,5-6 (short) >> JP18 : 1-2 (short) >> JP19 : 1-2,3-4,5-6 (short) >> JP20 : (none) >> JP21 : (none) >> JP22 : 1-2 (short) >> >> >> B) AMD 486DX4 (SV8B) >> Intel 486DX4 (&EW) >> >> JP16 : 1-2,4-5 (short) >> JP17 : 1-2,5-6 (short) >> JP18 : 1-2 (short) >> JP19 : 1-2,3-4,5-6 (short) >> JP20 : (none) >> JP21 : 3-4 (short) >> JP22 : 1-2,4-5 (short) >> >> C) Cyrix 486DX4 >> >> JP16 : 2-3,5-6 (short) >> JP17 : 2-3,5-6 (short) >> JP18 : 2-3 (short) >> JP19 : 2-3 (short) >> JP20 : 2-3 (short) >> JP21 : 1-2 (short) >> JP22 : 1-2 (short) >> >> D) AMD 486DX4 (NV8T) >> AMD 486DXL4 >> >> JP16 : 1-2,5-6 (short) >> JP17 : 4-5 (short) >> JP18 : 4-5 (short) >> JP19 : (none) >> JP20 : 4-5 (short) >> JP21 : (2-3 short : 2x CLK, 2-3 open : 3x CLK) >> JP22 : 1-2, 4-5 (short) >> >> >> --David >> +----------------------------------------------------------------------------+ >> | David A. Ranch - ISDN/Cablemodems/PCs dranch@ecst.csuchico.edu | >> +---- ----+ >> +----- For more detailed info, check out: www.ecst.csuchico.edu/~dranch -----+ >> >> >> > > > John and Nellena Difilippo difilipp@whidbey.net From difilipp@whidbey.net Thu Mar 28 21:42:17 1996 Received: from islander.whidbey.net (root@islander.whidbey.net [204.57.140.1]) by spam.ecst.csuchico.edu (8.7.3/8.7.3) with ESMTP id VAA01554 for ; Thu, 28 Mar 1996 21:42:15 -0800 (PST) Received: from difilippo (asn30.whidbey.net [204.57.140.100]) by islander.whidbey.net (8.7.5/8.7.3) with SMTP id VAA25473 for ; Thu, 28 Mar 1996 21:37:49 -0800 (PST) Message-Id: <2.2.32.19960329054207.0067ad2c@whidbey.net> X-Sender: difilipp@whidbey.net X-Mailer: Windows Eudora Pro Version 2.2 (32) Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Thu, 28 Mar 1996 21:42:07 -0800 To: "David A. Ranch" From: John and Nellena Difilippo Subject: Re: Setting for AMD 5x86-133 & GX4 : I've got the processor now! (fwd) Status: RO X-Status: Correction to JP29 GX4 rev 2.0 Award 4.50G version 0306 Promise EIDE 2300 Plus controller Diamond Stealth 64 Video series 2000 (2 Meg Dram) 16 Meg Ram (70 nsec) The settings I came up with are based on the AMD Enhanced 120's. The cpu's pin assignments are basically the same, the only difference seems to be what the the state of the input does to the chip. (clock multiplication and cache settings). I'm confident I've enabled L1 write back cache. These settings are still preliminary as I'm still waiting for the latest bios, but at present I'm running overclocked at 160Mhz and this $70 dollar chip is outperforming the $270 Pentium Overdrive 83 Mhz chip in every category including CPU 16 and 32 mark. (using winbench 96). I'm not sure it will last at 160Mhz, but the replacement of it won't be that expensive if it does give in. I know the next one will always work at 133Mhz. If these settings can be changed, I'm welcome to suggestions as I don't claim to have all of the answers. They are proven on my system. Bios settings: Auto Config Disabled Fast Reset Emul Enabled AT Bus Clock 1/4 CLKIN(1/5 for 160 Mhz) DRAM Speed Fastest DRAM Write WS 0WS DRAM Write CAS 1T DRAM Write Burst Enabled Slow Refresh Disabled Hidden Refresh Enabled L2 Cache Scheme Write Back L1 Cache Scheme Write Back Cache Burst Read 1T Cache Write Cycle 2T Video Shadow Non-cache Memory Hole Disabled Fast Reset Latency 2 usec Latch Local Bus T2 Local Bus Ready Transparent for 133Mhz Sychronize for 160Mhz Jumper Settings: JP5 1+2 short JP6 1+2 short JP11 2+3 short JP16 1+2, 4+5 short JP17 1+2, 5+6 short JP18 1+2 short JP19 2+3, 4+5 short JP20 1+2 short for 4x clock, All open for 3X clock JP21 3+4 short (should be write back cache), Open for write thru JP22 1+2, 4+5 short JP23, JP24, JP25 33Mhz for 133 operation, 40Mhz for 160 operation JP26 2+3 short JP28 1+2 short for 133Mhz, 2+3 short for 160Mhz JP29 1+2 short for 133Mhz, 2+3 short for 160Mhz JP32 1+2 short Just let me know if this worked for you. ASUS can give me a job! difilipp@whidbey.net At 12:30 PM 3/28/96 -0800, you wrote: > >Hey John, > >I'm in the process of upgrading my BIOS and getting the 133Mhz processor >to run at 160Mhz. Here is my current thread to two other people about the >process. Once I have ALL the details, I plan on making a master GX4 >upgrade document and posting it on the ASUS newsgroup. > >Could you give the technical specs a once-over and let me know if you >dis-agree with anything? > >I greatly appreciate it and I'll talk to you soon! > > >--David > +----------------------------------------------------------------------------+ > | David A. Ranch - ISDN/Cablemodems/PCs dranch@ecst.csuchico.edu | > +---- ----+ > +----- For more detailed info, check out: www.ecst.csuchico.edu/~dranch -----+ > >---------- Forwarded message ---------- >Date: Thu, 28 Mar 1996 14:27:01 -0500 (EST) >From: Roman Shapiro >To: "David A. Ranch" >Subject: Re: Setting for AMD 5x86-133 & GX4 : I've got the processor now! > >Here's something I read in a post recently: > >++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ >Hi >I've seen numerous requests for information on how to set the jumpers >correctly for this combination. I received two of these processors this >weekend, and managed to get them going using the following jumper >settings: > >JP16: 1-2 4-5 >JP17: 1-2 5-6 >JP18: 1-2 >JP19: 1-2 3-4 5-6 >JP20: 1-2 >JP21: 2-3 >JP22: 1-2 4-5 > >If you leave JP20 open the CPU runs in clock trebling mode instead of >clock quadrupling mode. You could experiment and try to get it to work >as a 3*50MHz system. Both CPUs ran flawlessly at both 133MHz (4*33MHz) >and 160MHz (4*40MHz). The only problem I did encounter was that QEMM >7.5 hung when it started, no matter what the speed of the CPU was. >I think it has something to do with QEMM identifying the CPU incorrectly. >Cheers >Graham Inggs >+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ > > > >On Thu, 28 Mar 1996, David A. Ranch wrote: > >> >> Hey Roman, >> >> Welp, I broke down and bought a 133Mhz processor. I'm in the process of >> burning in the v401 BIOS and I would to pick your brain about what the >> jumper settings are for the GX4 motherboard. Again, thanks for the help >> and I plan on posting all my results to the ASUS newsgroup. >> >> (from you earilier) >> >> > If you have the manual, I can tell you which entry to use (like AMD >> > Enhanced). >> >> Ok, the first thing is.. this processor is called the 5x86-133. The Asus >> WWW site keeps mentioning two AMD DX4 processors: SV8B and the NV8T. >> I'm under the impression that this 133Mhz processor is neither. Can you >> clear this up for me? Ok.. for the processor stuff, check out the bottom >> of this e-mail since I'll put all the differentGx4 processor types there >> and I'll include the readme file that came with the bios upgrade just in >> case. >> >The SV8 and SV8 stuff only applied to the 80-120 Mhz processors. The 133 >has and ADZ somewhere I think. > >> > With the chip running at 4x40, I have the speed to fastest and fewest >> > wait states (I'm pretty sure). But this is something that you have to >> > tweak in the end by yourself. At 4x33, you should be able to set things >> > to max speed. >> >> Great.. I'm running an AMD Dx/2 80 5V now and I can only run it at FASTER. >> I still might not be able to get my system to run at fastest but bah.. >> thats ok. Two question(s): >> >> 1) ASUS says that you MUST run the L1 cache in >> Write-Through mode instead of the preferable Write-Back mode if >> you have a bus-mastering card like a SCSI card (which I do). >> Have you had to deal with this or it doesn't apply for your >> situation? >> >Not sure. I use write-back and have everything except for a SCSI card. > >> 2) What settings are you using for your VLBus? >> >> - VESA delay ENABLED? (ASUS default is enabled) >> - <33Mhz & 0ws (This is what I'm running now) >> or >> >33Mhz & 1ws >> >At 40Mhz, I have the VESA delay enabled. It may work without it at 33Mhz. > >> >> OK.. lastly,,from the manual as you requested, here are the different CPU >> jumper settings. Could you note to me which I need to run the processor >> at both 133Mhz and 160Mhz for our GX4 motherboards? >> > >The same setting should allow you to run at 133 and 160- you just have to >change the bus speed (from 33 to 40). Compare the settings from the >beginning of this message to the differenct cpu setting from the manual. >I think it's closest to the AMD486DXL4 - 2x clock- you may not have to >change the jumpers much. :) There may be one jumper different- the one >that enables internal write-back. > >Also, be warned that at 160Mhz (4x40), my GX4 boots up and says 150Mhz. > >Good luck. >-Roman > >> --- >> >> probably no - SL Enhanced 486SX & SX2 or Non-SL 486 & SX2 >> >> probably no - SL Enhanced (non-SL) 486DX, DX2, 487SX & ODP; 486Dx/4 @ >> 3x clock >> >> probably no - 486DX4 - 2.5x clock >> >> probably no - 486DX4 - 2x clock (BIOS reports AM486DXL4 3x) >> >> possibly - P24D Write-Back L1 cache >> JP21 Short 2&3 - Write-Through L1 cache >> >> possibly - P24CT - Write-back L1 cache >> JP21 Short 2&3 - Write-Through L1 cache >> >> possibly - P24T - Write-back L1 cache >> JP21 Short 2&3 - Write-Through L1 cache >> 237-pin SL ODP >> >> possibly 120Mhz - 486DX4 ODP (OverDrive Processor?) - 3x clock >> >> possibly - 486DX4 ODP (OverDrive Processor?) - 2.5 clock >> >> doubtful - 486DX4 ODP (OverDrive Processor?) - 2.0 clock >> >> doubtful - AMD486SXL, AMD486SX2L >> >> doubtful - AMD486DXl, AMD486DX2L >> >> possible 120Mhz - AMD486DXL4 - 3x clock >> >> This is what I - AMD486DXL4 - 2x clock >> run my Dx/2 80 >> at >> >> doubtful - Cyrix 486DX, DX2 >> >> doubtful - Cyrix 486DX2-V >> >> doubtful - Cyrix DX5 - 3x clock >> >> doubtful - Cyrix DX5 - 2x clock >> >> >> ======= >> >> >> This is the text file that came with the BIOS upgrade image. I'm >> including this just in case it helps. >> >> ---- >> >> BIOS file : SV2G0401.AWD >> BIOS Ver. : 0401 (#401A0-0401) >> Related Mainboard : VL/I-486SV2G(X4) Rev. 2.0 or above >> Related Chipset : SiS 85C471 chipset >> >> >> Description : >> >> 1. To support the following CPUs : >> a.) AMD 486DX4 (SV8B) : with green & L1 cache write-back. >> b.) Intel 486DX4 (&EW) : with green & L1 cache write-back. >> c.) Cyrix 486DX4 : with green & L1 cache write-back. >> d.) Cyrix 5x86 (M1sc) : with Pentium code inside. >> e.) AMD 486DX4 (NV8T) : non-green & L1 cache write-through. >> >> The new CPUs jumper setting as the follow : >> >> A) Cyrix 486DX4 P/O >> Cyrix 5x86 (M1sc) >> >> JP16 : 1-2,5-6 (short) >> JP17 : 1-2,5-6 (short) >> JP18 : 1-2 (short) >> JP19 : 1-2,3-4,5-6 (short) >> JP20 : (none) >> JP21 : (none) >> JP22 : 1-2 (short) >> >> >> B) AMD 486DX4 (SV8B) >> Intel 486DX4 (&EW) >> >> JP16 : 1-2,4-5 (short) >> JP17 : 1-2,5-6 (short) >> JP18 : 1-2 (short) >> JP19 : 1-2,3-4,5-6 (short) >> JP20 : (none) >> JP21 : 3-4 (short) >> JP22 : 1-2,4-5 (short) >> >> C) Cyrix 486DX4 >> >> JP16 : 2-3,5-6 (short) >> JP17 : 2-3,5-6 (short) >> JP18 : 2-3 (short) >> JP19 : 2-3 (short) >> JP20 : 2-3 (short) >> JP21 : 1-2 (short) >> JP22 : 1-2 (short) >> >> D) AMD 486DX4 (NV8T) >> AMD 486DXL4 >> >> JP16 : 1-2,5-6 (short) >> JP17 : 4-5 (short) >> JP18 : 4-5 (short) >> JP19 : (none) >> JP20 : 4-5 (short) >> JP21 : (2-3 short : 2x CLK, 2-3 open : 3x CLK) >> JP22 : 1-2, 4-5 (short) >> >> >> --David >> +----------------------------------------------------------------------------+ >> | David A. Ranch - ISDN/Cablemodems/PCs dranch@ecst.csuchico.edu | >> +---- ----+ >> +----- For more detailed info, check out: www.ecst.csuchico.edu/~dranch -----+ >> >> >> > > > John and Nellena Difilippo difilipp@whidbey.net From ginggs@pluto.iafrica.com Fri Mar 29 08:20:12 1996 Received: from pluto.iafrica.com (root@pluto.iafrica.com [192.96.88.137]) by spam.ecst.csuchico.edu (8.7.3/8.7.3) with SMTP id IAA08688 for ; Fri, 29 Mar 1996 08:19:59 -0800 (PST) Received: from pluto.iafrica.com by pluto.iafrica.com with smtp (Smail3.1.29.1 #20) id m0u2gt9-0007LrC; Fri, 29 Mar 96 18:18 GMT+0200 Message-Id: Comments: Authenticated sender is From: "Graham Inggs" Organization: Internet Africa To: dranch@ecst.csuchico.edu Date: Fri, 29 Mar 1996 18:19:46 +0200 Subject: Re: Setting for AMD 5x86-133 & GX4 : I've got the processor Reply-to: ginggs@iafrica.com Priority: normal X-mailer: Pegasus Mail for Windows (v2.23) Status: RO X-Status: Hi David I installed two of these CPUs last weekend, one went into a system with a VLB SCSI controller and is being run at 160MHz, the other went into a system with a VLB Enhanced IDE controller and is being run at 133MHz. Both systems have ASUS GX4 motherboards. > > > 1) ASUS says that you MUST run the L1 cache in > > > Write-Through mode instead of the preferable Write-Back mode if > > > you have a bus-mastering card like a SCSI card (which I do). > > > I currently have an Adaptec 2842 Bus-mastering VLB SCSI > > > controller and I bet it will have possible problems. > > > > > > Have you had to deal with this or it doesn't apply for your > > > situation? > > > > > > (roman) > > Not sure. I use write-back and have everything except for a SCSI card. I couldn't get write-back to work with *either* system, I don't know what I did wrong there, could have been a CMOS setting or something. > > > 2) What settings are you using for your VLBus? > > > > > > - VESA delay ENABLED? (ASUS default is enabled) > > > - <33Mhz & 0ws (This is what I'm running now) > > > or > > > >33Mhz & 1ws > > > > > > > At 40Mhz, I have the VESA delay enabled. It may work without it at 33Mhz. On the 133Mhz system I have <33MHz, 0ws, and on the 160MHz system I have >33MHz, 1ws. > > Also, be warned that at 160Mhz (4x40), my GX4 boots up and says 150Mhz. Mine does the same, and for the 133MHz system it says 120MHz. We'll have to wait for BIOS 0402 (if there will be one) to fix that, I don't think its serious, I think ASUS just round off the calculated MHz value to one of the known speeds. > Also Gramm, did you upgrade your BIOS yourself? I have the 401 BIOS image > here and I'm going to burn it today but I'm not sure what format the image > is in. Is it in HEX, Object, etc? Yes, I also downloaded mine from ASUS, it is just a binary image. Just load it in as binary and dump it onto a blank 27C512. I would like to know how you get on. I am interested in comparing benchmarks. Norton SI 8, and VGABench. Also I'd like to get it running in Write Back mode. I have the data sheet for the 5x86, I'm planning on studying it and tracing out to what pins the jumpers go, and hopefully from that information we can work out the best jumper settings. I have a suspicion it can go faster. Cheers Graham Inggs From difilipp@whidbey.net Sat Mar 30 06:08:48 1996 Received: from islander.whidbey.net (root@islander.whidbey.net [204.57.140.1]) by spam.ecst.csuchico.edu (8.7.3/8.7.3) with ESMTP id GAA17903 for ; Sat, 30 Mar 1996 06:08:47 -0800 (PST) Received: from difilippo (asn20.whidbey.net [204.57.140.220]) by islander.whidbey.net (8.7.5/8.7.3) with SMTP id GAA10242 for ; Sat, 30 Mar 1996 06:04:22 -0800 (PST) Message-Id: <2.2.32.19960330140844.0068a20c@whidbey.net> X-Sender: difilipp@whidbey.net X-Mailer: Windows Eudora Pro Version 2.2 (32) Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Sat, 30 Mar 1996 06:08:44 -0800 To: "David A. Ranch" From: John and Nellena Difilippo Subject: Re: Setting for AMD 5x86-133 & GX4 : I've got the processor now! (fwd) Status: RO X-Status: A Thanx for the reply David, My system works all the way up to 1/3 clkin, at that point the CD Rom is no longer recognized at Win 95 Bootup. I stayed at 1/5 for reliability. I'll boost it to 1/4 and for a while to ensure theres reliability and a performance increase. I don't understand what you mean for the video cache. My system runs slower with cache enabled. As for the synchronize/transparent setting, at 133Mhz I'm able to run transparent. This setting allows the system to exchange data with the local bus immediately, vice waiting for the next clock pulse as in synchronize. My video, probably due to some other misadjustment in my setup, freezes while benching when the setting is at transparent at 160Mhz. Tranparent at 133Mhz increases performance. Synchronize at 160Mhz ensures reliabilty in my system. Look for a difference using both settings at 160Mhz and let me know what you came up with. If it works for you, I've got to dig a little deeper and find out what I missed. I'm still playing around so I'll keep intouch. Did you get my latest e-mail on the 29th? John At 12:01 AM 3/30/96 -0800, you wrote: > >Hey john, > >Ok.. I'm in the process of benching my Dx2-80 w/ the 304BIOS to get a >performance increase percentage. I'll get you my full results this >weekend. Until then, I wanted to point out a few things I noticed in >your setup: > >> AT Bus Clock 1/4 CLKIN(1/5 for 160 Mhz) > >Run you ISA bus at 10Mhz. (1/4 for 40Mhz bus) This speeds up some DOS >and ISA-related things. Only thing I found is that Iomegga Ditto3200 tape >drives don't like the higher speed. > >> Video Shadow Non-cache > >Enabling this creates a SIGNIFICANT speed performance in DOS and some >aspects of Windows. ENABLE THIS! > >> Local Bus Ready Transparent for 133Mhz >> Sychronize for 160Mhz > >Why the difference? I could never find a difference in performance. > > >Ok.. back to benching.. (man I hate waiting for all this!) > > >--David > +----------------------------------------------------------------------------+ > | David A. Ranch - ISDN/Cablemodems/PCs dranch@ecst.csuchico.edu | > +---- ----+ > +----- For more detailed info, check out: www.ecst.csuchico.edu/~dranch -----+ > > > John and Nellena Difilippo difilipp@whidbey.net From difilipp@whidbey.net Sat Mar 30 09:00:20 1996 Received: from islander.whidbey.net (root@islander.whidbey.net [204.57.140.1]) by spam.ecst.csuchico.edu (8.7.3/8.7.3) with ESMTP id JAA20981 for ; Sat, 30 Mar 1996 09:00:19 -0800 (PST) Received: from difilippo (asn108.whidbey.net [204.57.140.178]) by islander.whidbey.net (8.7.5/8.7.3) with SMTP id IAA14950 for ; Sat, 30 Mar 1996 08:55:55 -0800 (PST) Message-Id: <2.2.32.19960330170012.0067d1d4@whidbey.net> X-Sender: difilipp@whidbey.net X-Mailer: Windows Eudora Pro Version 2.2 (32) Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Sat, 30 Mar 1996 09:00:12 -0800 To: "David A. Ranch" From: John and Nellena Difilippo Subject: Re: Setting for AMD 5x86-133 & GX4 : I've got the processor now! (fwd) Status: RO X-Status: David, I tried 1/4 clkin again and once again showed a degradation in CPU 16 and 32 marks. Increases with 1/5 clkin. I read somewhere that the 133 works optimally with an 8 Mhz ISA bus speed and with a 33Mhz input. If I find it again I will get it to you. I also went back to transparent at 160 and video still froze during benchmark. Suspect it's just my older bios and will retest it tuesday when the new one gets here. All other settings seem to be optimal as each time I change them, performance somewhere gets worse or the system doesn't boot. I also found that my power saving features don't work. I'm sure it's a jumper setting for the chip and will let you know what I find out. Might be til after I get the new bios to see if it still doesn't work. Keep in touch, I think we're making good progress! John John, Thanx for the reply David, My system works all the way up to 1/3 clkin, at that point the CD Rom is no longer recognized at Win 95 Bootup. I stayed at 1/5 for reliability. I'll boost it to 1/4 and for a while to ensure theres reliability and a performance increase. I don't understand what you mean for the video cache. My system runs slower with cache enabled. As for the synchronize/transparent setting, at 133Mhz I'm able to run transparent. This setting allows the system to exchange data with the local bus immediately, vice waiting for the next clock pulse as in synchronize. My video, probably due to some other misadjustment in my setup, freezes while benching when the setting is at transparent at 160Mhz. Tranparent at 133Mhz increases performance. Synchronize at 160Mhz ensures reliabilty in my system. Look for a difference using both settings at 160Mhz and let me know what you came up with. If it works for you, I've got to dig a little deeper and find out what I missed. I'm still playing around so I'll keep intouch. Did you get my latest e-mail on the 29th? John At 12:01 AM 3/30/96 -0800, you wrote: > >Hey john, > >Ok.. I'm in the process of benching my Dx2-80 w/ the 304BIOS to get a >performance increase percentage. I'll get you my full results this >weekend. Until then, I wanted to point out a few things I noticed in >your setup: > >> AT Bus Clock 1/4 CLKIN(1/5 for 160 Mhz) > >Run you ISA bus at 10Mhz. (1/4 for 40Mhz bus) This speeds up some DOS >and ISA-related things. Only thing I found is that Iomegga Ditto3200 tape >drives don't like the higher speed. > >> Video Shadow Non-cache > >Enabling this creates a SIGNIFICANT speed performance in DOS and some >aspects of Windows. ENABLE THIS! > >> Local Bus Ready Transparent for 133Mhz >> Sychronize for 160Mhz > >Why the difference? I could never find a difference in performance. > > >Ok.. back to benching.. (man I hate waiting for all this!) > > >--David > +----------------------------------------------------------------------------+ > | David A. Ranch - ISDN/Cablemodems/PCs dranch@ecst.csuchico.edu | > +---- ----+ > +----- For more detailed info, check out: www.ecst.csuchico.edu/~dranch -----+ > > > John and Nellena Difilippo difilipp@whidbey.net From difilipp@whidbey.net Sun Mar 31 15:44:30 1996 Received: from islander.whidbey.net (root@islander.whidbey.net [204.57.140.1]) by spam.ecst.csuchico.edu (8.7.3/8.7.3) with ESMTP id PAA16592 for ; Sun, 31 Mar 1996 15:44:29 -0800 (PST) Received: from difilipp (asn44.whidbey.net [204.57.140.114]) by islander.whidbey.net (8.7.5/8.7.3) with SMTP id PAA17076 for ; Sun, 31 Mar 1996 15:39:51 -0800 (PST) Message-Id: <2.2.32.19960331234420.007637f4@whidbey.net> X-Sender: difilipp@whidbey.net X-Mailer: Windows Eudora Pro Version 2.2 (32) Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Sun, 31 Mar 1996 15:44:20 -0800 To: "David A. Ranch" From: John and Nellena Difilippo Subject: Re: Setting for AMD 5x86-133 & GX4 : I've got the processor now! (fwd) Status: RO X-Status: A Found it, will let you know what I come up with when I'm done downloading it. John John and Nellena Difilippo difilipp@whidbey.net From difilipp@whidbey.net Tue Apr 2 18:52:44 1996 Received: from islander.whidbey.net (root@islander.whidbey.net [204.57.140.1]) by spam.ecst.csuchico.edu (8.7.3/8.7.3) with ESMTP id SAA06834 for ; Tue, 2 Apr 1996 18:52:42 -0800 (PST) Received: from difilipp (asn81.whidbey.net [204.57.140.151]) by islander.whidbey.net (8.7.5/8.7.3) with SMTP id SAA28004 for ; Tue, 2 Apr 1996 18:48:03 -0800 (PST) Message-Id: <2.2.32.19960403025225.006958a8@whidbey.net> X-Sender: difilipp@whidbey.net X-Mailer: Windows Eudora Pro Version 2.2 (32) Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Tue, 02 Apr 1996 18:52:25 -0800 To: "David A. Ranch" From: John and Nellena Difilippo Subject: Re: I've got SOMETHING running now(GX4)... VERY weird. Status: RO X-Status: David, I'm running the 0401 bios with the same performance I previously had. I also have power saving features, which I didn't have before. Haven't gone into the settings yet to see how its different or what I can do. John. At 04:21 PM 3/31/96 -0800, you wrote: > >Well, strike 1: a 4x33Mhz still reported a 120Mhz for me > >So, I played around and this is the best yet (TOTALLY different from >your setup). This is basicallt CONFIRMING that there are sub-versions to >the GX4 v2.0 motherboard. I received one of the very first GX4 boards >available and had to haggle with the jumpers like this to get the DX/2 >80Mhz to run well. This is what I'm tring currently: > >JP5: 1+2 P24D, P24T, P24CT, AMD(s) X PLUS hardware trap >JP6: 1+2 > >JP11: 2+3 Intel processor compatible > > Close to a 485DX4 - 2 x clock setting >JP16: 1+2 5+6 CPU setting >JP17: 1+2 5+6 CPU setting >JP18: 1+2 5+6 CPU setting >JP19: 1+2 CPU setting >JP20: 1+2 4x CLOCK >JP21: 2+3 L1 WT cache setting (DMA conflict fix) >JP22: 1+2 3+4 CPU setting > >JP23: 2+3 40Mhz clock >JP24: 2+3 >JP25: 1+2 > >JP26: 1+2 VLB Delay > >JP28: 1+2 VLB 0ws >JP29: 1+2 VLB <33Mhz > >JP32: 1+2 Intel DX4 > > > > >--David > +----------------------------------------------------------------------------+ > | David A. Ranch - ISDN/Cablemodems/PCs dranch@ecst.csuchico.edu | > +---- ----+ > +----- For more detailed info, check out: www.ecst.csuchico.edu/~dranch -----+ > > > John and Nellena Difilippo difilipp@whidbey.net From ginggs@pluto.iafrica.com Wed Apr 3 00:18:53 1996 Received: from hairball.ecst.csuchico.edu (root@hairball.ecst.csuchico.edu [132.241.1.108]) by spam.ecst.csuchico.edu (8.7.3/8.7.3) with SMTP id AAA16518 for ; Wed, 3 Apr 1996 00:18:52 -0800 (PST) Received: from pluto.iafrica.com (pluto.iafrica.com [192.96.88.137]) by hairball.ecst.csuchico.edu (8.6.12/8.6.12) with SMTP id AAA28415 for ; Wed, 3 Apr 1996 00:18:46 -0800 Received: from pluto.iafrica.com by pluto.iafrica.com with smtp (Smail3.1.29.1 #20) id m0u4Ndn-0007McC; Wed, 3 Apr 96 10:10 GMT+0200 Message-Id: Comments: Authenticated sender is From: "Graham Inggs" Organization: Internet Africa To: dranch@ecst.csuchico.edu Date: Wed, 3 Apr 1996 10:11:05 +0200 Subject: Re: Setting for AMD 5x86-133 & GX4 : I've got the processor Reply-to: ginggs@iafrica.com Priority: normal X-mailer: Pegasus Mail for Windows (v2.23) Status: RO X-Status: Hi David > Ok, was than an Adaptec, BusLogic, etc? It should matter since they both > (SCSI and EIDE) are probably bus-mastering cards. I didn't get a chance > to install the card today but I will tonight. I also plan on benchmarking I have a Tekram DC380 SCSI adaptor, in the manual for this card is specifically says it does not support CPUs with write back cache. > > I couldn't get write-back to work with *either* system, I don't know > > what I did wrong there, could have been a CMOS setting or something. > When you say couldn't get it to work, what do you mean? Its documented > that there is something incompatible with the GX4, Write-Back, and a > bus-mastering VLB card. Hmmmm... I'll play with this. Well, I tried setting the jumper to get Write Back, but neither system would boot up, I thought that at least the IDE system would work. > Did the 160Mhz machine puke when you set it to <33Mhz and 0ws? I'm > currently running my DX/2-80 (5V) at that but I did have to set my RAM > setting to FASTER. I wonder which is faster combination? I also have a DX2/80 (actually an overclocked Intel DX2/66) and I had starnge problems running that when it was set to <33MHz, 0ws. Occasionally windows would get corrupted when moved around, or opened and closed, so I set it to >33Mhz, 1ws, and never had a problem since. I was then able to set my DRAM access to FASTEST, and my SRAM to 2-1-1-1. > Did you notice any cosmetic, new features, etc in the new BIOS code? It depends which version of the bios you had. > Ok.. I'll run these two but I recommend a few others: > -Qbench by Quantum (ftp.quantum.com): great HD performance/throughput > -Bench32 for Windows95 > -PCBench (i'm not sure what version it is but its very comprehensive) I'll have a look out for them, and get back to you. I'm running Novell DOS 7, by the way, not Win95. > Also.. what other hardware do you have in these boxes: OK, the 160MHz system has the following: 16MB 60ns DRAM 256K 15ns SRAM Tekram DC380 VL SCSI adaptor Cirrus Logic 5428 VL SVGA (2MB) Sound Blaster 32 AWE Intel EtherExpress 16 bit NIC Microcom 28.8K internal modem Non-Branded Serial / Parallel Game card Conner CFP 1060MB SCSI-2 hard drive Conner 420MB Tape Drive Creative Labs Hex Speed CDROM 1.44MB stiffy drive The 133MHz system has : 16MB 60ns DRAM 128K 15ns SRAM Acer Enhanced IDE / Cirrus Logic 5430 VL Combo Card IBM 1.7GB IDE hard drive 1.44MB stiffy drive 1.2MB floppy drive > ANyway.... its great to have another non-fearful computer hacker out there > on the ASUS side that actually and spell english and it comprehensive in > the e-mail dept! I'll get cracking on that tonight and I'll send you some > e-mails on my results, benchmarks, etc! Ditto! Cheers Graham From dranch@ecst.csuchico.edu Sun Mar 31 16:22:02 1996 Received: from corpse.ecst.csuchico.edu (root@corpse.ecst.csuchico.edu [132.241.5.10]) by spam.ecst.csuchico.edu (8.7.3/8.7.3) with SMTP id QAA17606 for ; Sun, 31 Mar 1996 16:22:01 -0800 (PST) Received: from localhost (dranch@localhost [127.0.0.1]) by corpse.ecst.csuchico.edu (8.6.12/8.6.12) with SMTP id QAA12125; Sun, 31 Mar 1996 16:21:56 -0800 Date: Sun, 31 Mar 1996 16:21:55 -0800 (PST) From: "David A. Ranch" To: John and Nellena Difilippo cc: "David A. Ranch" Subject: I've got SOMETHING running now(GX4)... VERY weird. In-Reply-To: <2.2.32.19960331234420.007637f4@whidbey.net> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Status: RO X-Status: Well, strike 1: a 4x33Mhz still reported a 120Mhz for me So, I played around and this is the best yet (TOTALLY different from your setup). This is basicallt CONFIRMING that there are sub-versions to the GX4 v2.0 motherboard. I received one of the very first GX4 boards available and had to haggle with the jumpers like this to get the DX/2 80Mhz to run well. This is what I'm tring currently: JP5: 1+2 P24D, P24T, P24CT, AMD(s) X PLUS hardware trap JP6: 1+2 JP11: 2+3 Intel processor compatible Close to a 485DX4 - 2 x clock setting JP16: 1+2 5+6 CPU setting JP17: 1+2 5+6 CPU setting JP18: 1+2 5+6 CPU setting JP19: 1+2 CPU setting JP20: 1+2 4x CLOCK JP21: 2+3 L1 WT cache setting (DMA conflict fix) JP22: 1+2 3+4 CPU setting JP23: 2+3 40Mhz clock JP24: 2+3 JP25: 1+2 JP26: 1+2 VLB Delay JP28: 1+2 VLB 0ws JP29: 1+2 VLB <33Mhz JP32: 1+2 Intel DX4 --David +----------------------------------------------------------------------------+ | David A. Ranch - ISDN/Cablemodems/PCs dranch@ecst.csuchico.edu | +---- ----+ +----- For more detailed info, check out: www.ecst.csuchico.edu/~dranch -----+ From difilipp@whidbey.net Sun Mar 31 19:02:28 1996 Received: from islander.whidbey.net (root@islander.whidbey.net [204.57.140.1]) by spam.ecst.csuchico.edu (8.7.3/8.7.3) with ESMTP id TAA21663 for ; Sun, 31 Mar 1996 19:02:25 -0800 (PST) Received: from difilipp (asn111.whidbey.net [204.57.140.181]) by islander.whidbey.net (8.7.5/8.7.3) with SMTP id SAA25368 for ; Sun, 31 Mar 1996 18:58:00 -0800 (PST) Message-Id: <2.2.32.19960401030221.0068d8bc@whidbey.net> X-Sender: difilipp@whidbey.net X-Mailer: Windows Eudora Pro Version 2.2 (32) Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Sun, 31 Mar 1996 19:02:21 -0800 To: "David A. Ranch" From: John and Nellena Difilippo Subject: Re: cachechk (fwd) Status: RO X-Status: Hello David, I received your file and ran it but am not sure what the results mean. All test ran the same and the bottom looked something like this: 3.9 clks 8.6 clks 13 clks 333% 150% 100% John At 06:32 PM 3/31/96 -0800, you wrote: >Content-Type: TEXT/PLAIN; CHARSET=us-ascii >Content-ID: > > >Hey John, > >Run this for memory throughput tests... I'll get you my final benchmarks >once I'm done! > >THis is a MIME encapsulated file.. I hope your mail reader can handle >MIME. If not, let me know. > >--David > +----------------------------------------------------------------------------+ > | David A. Ranch - ISDN/Cablemodems/PCs dranch@ecst.csuchico.edu | > +---- ----+ > +----- For more detailed info, check out: www.ecst.csuchico.edu/~dranch -----+ > >---------- Forwarded message ---------- >Date: Sun, 31 Mar 1996 17:49:53 -0800 >From: Andre Hendarman >To: dranch@ecst.csuchico.edu >Subject: cachechk > >here's the latest cachechk v.4 > >enjoy, > >andre >Content-Type: APPLICATION/ZIP; NAME="CACHCHK4.ZIP"; X-MAC-TYPE=705A4950; X-MAC-CREATOR=705A4950 >Content-ID: >Content-Description: > >Attachment Converted: C:\Upload\CACHCHK4.zip >Content-Type: TEXT/PLAIN; CHARSET=us-ascii >Content-ID: >Content-Description: > >/---------------------------------------------------------\ >| Andre Hendarman | >| Vitesse Semiconductor, ATE Group | >| andre@ix.netcom.com | >| andre@vitsemi.com | >| | >| "GaAs is the future" | >\---------------------------------------------------------/ > John and Nellena Difilippo difilipp@whidbey.net From difilipp@whidbey.net Wed Apr 3 11:17:33 1996 Received: from islander.whidbey.net (root@islander.whidbey.net [204.57.140.1]) by spam.ecst.csuchico.edu (8.7.3/8.7.3) with ESMTP id LAA22177 for ; Wed, 3 Apr 1996 11:17:32 -0800 (PST) Received: from difilipp (asn137.whidbey.net [204.57.140.26]) by islander.whidbey.net (8.7.5/8.7.3) with SMTP id LAA08744 for ; Wed, 3 Apr 1996 11:12:54 -0800 (PST) Message-Id: <2.2.32.19960403191719.00688c9c@whidbey.net> X-Sender: difilipp@whidbey.net X-Mailer: Windows Eudora Pro Version 2.2 (32) Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Wed, 03 Apr 1996 11:17:19 -0800 To: "David A. Ranch" From: John and Nellena Difilippo Subject: Re: Setting for AMD 5x86-133 & GX4 : I've got the processor Status: RO X-Status: At 10:40 AM 4/3/96 -0800, you wrote: > >> I have a Tekram DC380 SCSI adaptor, in the manual for this card is >> specifically says it does not support CPUs with write back cache. > >Thats really weird. The only reason I can think of why they would say >that is because WB cache waits for idle peroids before they update main >memory. If the controller bypasses cache (which I would highly doubt), >that would hose the controller. > >Anyway.. I haven't gotten the chance to finish up on the docs for what I >did but it turns out that I had to use a RADICALLY different jumper >settings than you or John used. I had this sam problem when I upgraded to >my old DX2/80. Why? Dunno... I think I might have a pre/post-2.0 board >though its 2.0 silkscreened on the board. > >Anyway.. Have you run PCBench9 or WinBench96? I've noticed a disturbing >attribute. If the L2 cache is WB, I get better PcBench numbers (152 vs >128), if its WT, then I get better WinBench numbers (129 vs. 108)! I run >mostly in Win95 but when I play games like quake, etc.. do I have to >change CMOS? I DOUBT it.. hmmm... what do you think? > >Ps. I'm going to forward this to John since he's doing exactly what we >are doing and he might have some ideas too (hey john!) > >--David > +----------------------------------------------------------------------------+ > | David A. Ranch - ISDN/Cablemodems/PCs dranch@ecst.csuchico.edu | > +---- ----+ > +----- For more detailed info, check out: www.ecst.csuchico.edu/~dranch -----+ > > > Don't know about the scuzzy card. I definately get worst performance using write back L2 cache setting in windows. Didn't test dos. I have the 401 bios and all of my previous settings have turned out to be optimal. Every other change is for the worst, including the jumper settings. I have power saving features that I didnt have with the 306 bios so that means that the 401 has made this chip and my system more compatable. With the 401, I'm also able to vary the cpu modes,(write thru and write back) although I saw no difference withnthe 306. I get a drasticly worse cpu mark when I go to write thru in the cmos setup. I'm positive it's operating in write back mode as indicated by my benchmark prog, and to my surprise there are no glitches. Good luck! John and Nellena Difilippo difilipp@whidbey.net From ginggs@pluto.iafrica.com Wed Apr 3 13:54:10 1996 Received: from pluto.iafrica.com (root@pluto.iafrica.com [192.96.88.137]) by spam.ecst.csuchico.edu (8.7.3/8.7.3) with SMTP id NAA28547 for ; Wed, 3 Apr 1996 13:54:03 -0800 (PST) Received: from pluto.iafrica.com by pluto.iafrica.com with smtp (Smail3.1.29.1 #20) id m0u4aV0-0007MfC; Wed, 3 Apr 96 23:53 GMT+0200 Message-Id: Comments: Authenticated sender is From: "Graham Inggs" Organization: Internet Africa To: dranch@ecst.csuchico.edu Date: Wed, 3 Apr 1996 23:54:56 +0200 Subject: Re: Setting for AMD 5x86-133 & GX4 : I've got the processor Reply-to: ginggs@iafrica.com Priority: normal X-mailer: Pegasus Mail for Windows (v2.23) Status: RO X-Status: Hi David > > I have a Tekram DC380 SCSI adaptor, in the manual for this card is > > specifically says it does not support CPUs with write back cache. > Thats really weird. The only reason I can think of why they would say > that is because WB cache waits for idle peroids before they update main > memory. If the controller bypasses cache (which I would highly doubt), > that would hose the controller. The way I see it, the L2 cache is wired directly to the CPU, and not (as I think it should be) to the memory system. The method used for WB cache obviously assumes (incorrectly) that it is the only device that can be in control of the bus, and doesn't bother to check. > Anyway.. I haven't gotten the chance to finish up on the docs for what I > did but it turns out that I had to use a RADICALLY different jumper > settings than you or John used. I had this sam problem when I upgraded to > my old DX2/80. Why? Dunno... I think I might have a pre/post-2.0 board > though its 2.0 silkscreened on the board. Please let me know what your jumper settings are/were. > Anyway.. Have you run PCBench9 or WinBench96? I've noticed a disturbing > attribute. If the L2 cache is WB, I get better PcBench numbers (152 vs > 128), if its WT, then I get better WinBench numbers (129 vs. 108)! I run > mostly in Win95 but when I play games like quake, etc.. do I have to > change CMOS? I DOUBT it.. hmmm... what do you think? I've just noticed something similar, I am getting a faster benchmark with my L2 cache *disabled* than when it is enabled. At the moment it is WB, I will set it to WT and see if that helps at all. I am trying to download WinBench 96 right now so we can make direct comparisons. In case I haven't told you before, I am running Novell Dos 7, Windows 3.1, Win32S 1.30 and used to use QEMM 7.5 (which hasn't worked since I put in the new CPU). I have 256K (32K x 8) of 15ns L2 cache memory. Cheers Graham From dranch@ecst.csuchico.edu Wed Apr 10 13:21:55 1996 Received: from corpse.ecst.csuchico.edu (dranch@corpse.ecst.csuchico.edu [132.241.5.10]) by spam.ecst.csuchico.edu (8.7.3/8.7.3) with SMTP id NAA07014 for ; Wed, 10 Apr 1996 13:21:53 -0700 (PDT) Received: (from dranch@localhost) by corpse.ecst.csuchico.edu (8.6.12/8.6.12) id NAA06965; Wed, 10 Apr 1996 13:20:54 -0700 Date: Wed, 10 Apr 1996 13:20:54 -0700 From: "David A. Ranch" Message-Id: <199604102020.NAA06965@corpse.ecst.csuchico.edu> To: dranch@ecst.csuchico.edu, carl@bre.com, c.michalik@bre.com, ginggs@iafrica.com, difilipp@whidbey.net Newsgroups: alt.comp.periphs.mainboard.asus Subject: AMD 133(160Mhz) settings for odd ASUS GX4 motherboards w/ BENCHMARKS References: <4kdqms$4dq@brazos.bihs.net> <4kf239$b7o@charnel.ecst.csuchico.edu> Organization: California State University, Chico Status: RO X-Status: David A. Ranch wrote: >I had the EXACT same thing happen to me. I will get the new jumper >settings to you as soon as I get home. It took me -3- days to figure >these out!! #### Ok, here are the jumper / BIOS settings and quick benchmarks that I found for a GX4 v2.0 board and v401 of the Award BIOS: ASUS GX4 v2.0 board 16MB FPM DRAM 256K async cache Cirrus Logic 5429 VLB video w/ 2MB DRAM Adaptec 2842 SCSI-II-Fast Controller IBM UltraStar XP 2.2GB disk Gravis Ultrasound w/ 1MB DRAM Iomega Ditto 3200 tape drive JP5 : 1+2 Intel setting JP6 : 2+3 Intel setting JP11: 2+3 Intel 3.45v CPU JP16: 1+2,5+6 CPU setting JP17: 1+2,5+6 CPU setting JP18: 1+2,5+6 CPU setting JP19: 1+2 CPU setting JP20: 1+2 4x clock setting (open for 3x clock) JP21: 2+3 for Write Thru (3+4 for Write Back) Note: I had to use WT cache because I have an Adaptec 2842 busmastering controller. Try the WB setting first! JP22: 1+2,4+5 CPU setting (for 160Mhz) JP23: 2+3 JP24: 2+3 JP25: 1+2 JP26: 1+2 VLB delay JP28: 1+2 VLB 0ws JP29: 1+2 VLB <33Mhz ----- BIOS settings: Note: these are pushed so mileage will vary.. you must tweak with these to get a stable system Auto config : disbabled AT BUS Clock : 1/4 CLKIN (10Mhz AT BUS at 40Mhz system clock) DRAM Speed : Faster (My machine won't run stable at FASTEST) DRAM Write WS : 0ws DRAM Write CAS : 1T DRAM Write Burst : Enabled Slow Refresh : Enabled Hidden Refresh : Enabled L2 Cache scheme : WB - for DOS WT - for Windows (see below) L1 Cache scheme : WT (I had to use WT or my machine wouldn't boot) Cache Burst Read : 1T Cache Write Cycle : 2T Video Shadow : non-cachable Memory Hole : Disabled Fast Reset Latency: 2 uS Latch Local Bus : T2 Local Bus Ready : Syncronize --- Note: I'm not sure why but PCbench and DOS benchmarks run FASTER with the L2 cache sent to Auto (WT) caching. Any ideas WHY?!?!?! I've left my machine in WT mode since I use Windows95 most of the time. L2-Write Through PCBench9: CPU16: 128 Video: 5662 Disk : 224 Norton SI/95: 16.9 WinBench96: CPU16: 129 CPU32: 130 L2-Write Back PCBench9: CPU16: 152 Video: n/a Disk: n/a Norton SI/95: n/a WinBench96: CPU16: 120 CPU32: 116 Video: 5675 Disk: 223 -- +----------------------------------------------------------------------------+ | David A. Ranch - ISDN/Cablemodems/PCs dranch@ecst.csuchico.edu | +---- ----+ +----- For more detailed info, check out: www.ecst.csuchico.edu/~dranch -----+ From dranch@ecst.csuchico.edu Wed Apr 10 15:44:43 1996 Received: from pluto.iafrica.com (root@pluto.iafrica.com [192.96.88.137]) by spam.ecst.csuchico.edu (8.7.3/8.7.3) with SMTP id PAA21505 for ; Wed, 10 Apr 1996 15:44:39 -0700 (PDT) Received: from pluto.iafrica.com by pluto.iafrica.com with smtp (Smail3.1.29.1 #20) id m0u78cj-0007LCC; Thu, 11 Apr 96 00:44 GMT+0200 Message-Id: Comments: Authenticated sender is From: "Graham Inggs" Organization: Internet Africa To: "David A. Ranch" Date: Thu, 11 Apr 1996 00:43:29 +0200 Subject: Re: AMD 133(160Mhz) settings for odd ASUS GX4 motherboards w/ B Reply-to: ginggs@iafrica.com Priority: normal X-mailer: Pegasus Mail for Windows (v2.23) Status: RO X-Status: A Greetings David > JP16: 1+2,5+6 CPU setting > JP17: 1+2,5+6 CPU setting > JP18: 1+2,5+6 CPU setting > JP19: 1+2 CPU setting > JP20: 1+2 4x clock setting (open for 3x clock) > JP21: 2+3 for Write Thru (3+4 for Write Back) > Note: I had to use WT cache because I have an Adaptec 2842 > busmastering controller. Try the WB setting first! > JP22: 1+2,4+5 CPU setting How did you arrive at these CPU settings? What happens if you set it up the same way as I did? JP16: 1-2,4-5 JP17: 1-2,5-6 JP18: 1-2 JP19: 1-2,3-4,5-6 JP20: 1-2 JP21: 2-3 JP22: 1-2,4-5 This is the same as ASUS recommends for an AMD DX4 SV8B on a GX4 mainboard, and for a 5x86 on an SP3 mainboard. > AT BUS Clock : 1/4 CLKIN (10Mhz AT BUS at 40Mhz system clock) Mine is set at 1/5 CLKIN, does it help pushing the ISA bus a bit? > DRAM Speed : Faster (My machine won't run stable at FASTEST) Mine is on Fastest (my DRAM is 60ns, and my SRAM 15ns) >From past experience, even if you have 70ns DRAM, you can still set it to Fastest with a 40MHz bus clock, but you have to slow down if your SRAM is 20ns. > Video Shadow : non-cachable Mine is cacheable, do you know any more about this setting? When should it be enabled, and when not? > Latch Local Bus : T2 I don't know much about this one, but when I was setting up my DX4/100MHz as a 2 * 50MHz, I could only get the system to be stable when this was set to T3, so I left it that way. Maybe this could make your system work at the Fastest setting. > Local Bus Ready : Syncronize I don't know anything about this one either, but mine is the same. > Note: I'm not sure why but PCbench and DOS benchmarks run FASTER with the > L2 cache sent to Auto (WT) caching. Any ideas WHY?!?!?! I've > left my machine in WT mode since I use Windows95 most of the time. I've also had some weird benchmark results. Try disabling your L2 cache completely, and check your results. WinBench 96: GWM 96 CPU 16 CPU 32 L2-WB 5.36 94.9 94.1 L2-WT 5.43 94.9 94.1 L2-none 5.29 94.6 94.2 I'm using Windows 3.1 and Win32S 1.30, so I'm not sure if that is the cause of the lower scores. Either my L2 is not working at all, or something else strange is going on. I saw in the newsgroup that some other people have also found that the L2 cache doesn't make much of a difference. I've been thinking about this, and I know that with the older 486's (DX, DX2, DX4) that going from 0K to 128K L2 cache makes a big difference, but going from 128K to 256K is not noticeable at all. Now I've been wondering just how much of the 128K is actually being used during everyday working. One of the big differences between the 5x86 and older 486's is the 5x86 has 16K as opposed to 8K L1 cache. Now if the loops in the benchmark tests are >8K but <16K that would explain why disabling the L2 has little effect on the score. Otherwise, I either have a faulty CPU or I should look at trying out another motherboard. Cheers Graham From dranch@ecst.csuchico.edu Wed Apr 10 16:02:03 1996 Status: RO X-Status: A Received: from islander.whidbey.net (root@islander.whidbey.net [204.57.140.1]) by spam.ecst.csuchico.edu (8.7.3/8.7.3) with ESMTP id QAA23236 for ; Wed, 10 Apr 1996 16:02:02 -0700 (PDT) Received: from difilipp (asn101.whidbey.net [204.57.140.171]) by islander.whidbey.net (8.7.5/8.7.3) with SMTP id PAA24349 for ; Wed, 10 Apr 1996 15:57:24 -0700 (PDT) Message-Id: <2.2.32.19960410230156.00693b10@whidbey.net> X-Sender: difilipp@whidbey.net X-Mailer: Windows Eudora Pro Version 2.2 (32) Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Wed, 10 Apr 1996 16:01:56 -0700 To: "David A. Ranch" From: John and Nellena Difilippo Subject: Re: AMD 133(160Mhz) settings for odd ASUS GX4 motherboards w/ BENCHMARKS David, I've had two people reply that they have successfully configured their system at 160 and 133 respectivly. The systems have proven to have better performance with L2 set to write thru (auto), but both cases have had no problem with the write back setting for L1 cache. The person running 160 has cpu marks just below mine. Are you sure that you are going to be able to settle for write thru with the cpu marks at the 120's? Any thing I can do to help, let me know. John Date: Tue, 9 Apr 1996 23:03:29 -0400 X-Sender: simon@magmacom.com To: John and Nellena Difilippo From: Craig Simon Subject: Re: AMD 133 in GX4 success story Content-Length: 883 Hi, Just thought I would write to thank you for your information concerning the AMD-133. I have now successfully updated to this processor and overclocked it. It is running stabley at 160 Mhz. WinBench 96 CPUmark16 and CPUmark32 are both at 172. Comparable to your results. I updated my BIOS to 0401 before I started. This led to an interesting observation. I initially set the cache to write-back (the other option with this BIOS is auto) and this only gave me CPUmarks of 130-140. When I changed the cache settings to AUTO my CPUmarks jumped to 172. This seems to parallel your observation about setting the 0306 BOIS cache settings to write-through. Another minor note. At start-up the bios reports the clock speed as 120 or 150 for VLB settings of 33 and 40, respectively. This appears to be meaningless as winbench detects the CPU correctly. Thanks again, Craig X-Sender: jscole@telalink.net Date: Sat, 06 Apr 1996 20:34:17 -0600 To: John and Nellena Difilippo From: "J. S. Cole" Subject: Re: AMD 133 in GX4 overclocked to 160 Content-Length: 595 At 10:03 PM 4/4/96 -0800, you wrote: >Let me know if they work. > > Your settings seem to work. I can't seem to find a good program that will report true clock speed. One program reported 140mhz while the other reported 236 mhz and the BIOS shows 120 mhz. I haven't bumped to 160 yet but you and I have exactly the same cards (EIDE 2300 and Stealth 64) so I'll give it a shot later this week. What is the function of "Latch Local Bus" and "Local Bus Ready" ? I had mine set to T3 and Synchronize but changed it to T2 and Transparent based on your settings. I haven't noticed any difference. At 01:20 PM 4/10/96 -0700, you wrote: >David A. Ranch wrote: > >>I had the EXACT same thing happen to me. I will get the new jumper >>settings to you as soon as I get home. It took me -3- days to figure >>these out!! > >#### > >Ok, here are the jumper / BIOS settings and quick benchmarks that I found >for a GX4 v2.0 board and v401 of the Award BIOS: > > ASUS GX4 v2.0 board > 16MB FPM DRAM > 256K async cache > Cirrus Logic 5429 VLB video w/ 2MB DRAM > Adaptec 2842 SCSI-II-Fast Controller > IBM UltraStar XP 2.2GB disk > Gravis Ultrasound w/ 1MB DRAM > Iomega Ditto 3200 tape drive > > > >JP5 : 1+2 Intel setting >JP6 : 2+3 Intel setting > >JP11: 2+3 Intel 3.45v CPU > >JP16: 1+2,5+6 CPU setting >JP17: 1+2,5+6 CPU setting >JP18: 1+2,5+6 CPU setting >JP19: 1+2 CPU setting >JP20: 1+2 4x clock setting (open for 3x clock) >JP21: 2+3 for Write Thru (3+4 for Write Back) > Note: I had to use WT cache because I have an Adaptec 2842 > busmastering controller. Try the WB setting first! >JP22: 1+2,4+5 CPU setting > >(for 160Mhz) >JP23: 2+3 >JP24: 2+3 >JP25: 1+2 > >JP26: 1+2 VLB delay > >JP28: 1+2 VLB 0ws >JP29: 1+2 VLB <33Mhz > >----- > >BIOS settings: Note: these are pushed so mileage will vary.. you must > tweak with these to get a stable system > >Auto config : disbabled >AT BUS Clock : 1/4 CLKIN (10Mhz AT BUS at 40Mhz system clock) >DRAM Speed : Faster (My machine won't run stable at FASTEST) >DRAM Write WS : 0ws >DRAM Write CAS : 1T > >DRAM Write Burst : Enabled >Slow Refresh : Enabled >Hidden Refresh : Enabled > >L2 Cache scheme : WB - for DOS WT - for Windows (see below) >L1 Cache scheme : WT (I had to use WT or my machine wouldn't boot) >Cache Burst Read : 1T >Cache Write Cycle : 2T >Video Shadow : non-cachable >Memory Hole : Disabled > >Fast Reset Latency: 2 uS >Latch Local Bus : T2 >Local Bus Ready : Syncronize > > >--- >Note: I'm not sure why but PCbench and DOS benchmarks run FASTER with the > L2 cache sent to Auto (WT) caching. Any ideas WHY?!?!?! I've > left my machine in WT mode since I use Windows95 most of the time. > >L2-Write Through > > PCBench9: CPU16: 128 > Video: 5662 > Disk : 224 > > Norton SI/95: 16.9 > > WinBench96: CPU16: 129 > CPU32: 130 > > >L2-Write Back > > PCBench9: CPU16: 152 > Video: n/a > Disk: n/a > > Norton SI/95: n/a > > WinBench96: CPU16: 120 > CPU32: 116 > Video: 5675 > Disk: 223 >-- > +----------------------------------------------------------------------------+ > | David A. Ranch - ISDN/Cablemodems/PCs dranch@ecst.csuchico.edu | > +---- ----+ > +----- For more detailed info, check out: www.ecst.csuchico.edu/~dranch -----+ > > John and Nellena Difilippo difilipp@whidbey.net From difilipp@whidbey.net Fri Mar 29 20:19:46 1996 Received: from islander.whidbey.net (root@islander.whidbey.net [204.57.140.1]) by spam.ecst.csuchico.edu (8.7.3/8.7.3) with ESMTP id UAA18696 for ; Fri, 29 Mar 1996 20:19:37 -0800 (PST) Received: from difilippo (asn192.whidbey.net [204.57.140.81]) by islander.whidbey.net (8.7.5/8.7.3) with SMTP id UAA23778 for ; Fri, 29 Mar 1996 20:14:51 -0800 (PST) Message-Id: <2.2.32.19960330041912.006abd98@whidbey.net> X-Sender: difilipp@whidbey.net X-Mailer: Windows Eudora Pro Version 2.2 (32) Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Fri, 29 Mar 1996 20:19:12 -0800 To: "David A. Ranch" From: John and Nellena Difilippo Subject: Re: Setting for AMD 5x86-133 & GX4 : I've got the processor now! (fwd) Status: RO X-Status: Changed L2 write back to write thru in cmos setup. Gots an extra 40 point in CPU 16 and 32 mark. REQUIRED DISCLOSURE Benchmark Ziff-Davis' WinBench 96 Version 1.0 CDROM Controller Unknown CDROM Controller RAM ? KB CDROM Drive Unknown CDROM Software Cache CDFS Cache CDROM Software Cache Size 1088 KB CoProcessor On-Chip Display Adapter Diamond Stealth 64 Series (Diamond GT) Display DRAM ? KB Display Driver C:\WIN95\SYSTEM\DDOM95.DRV, Feb 28 10:50:50 1996, 89088 bytes, 4.02.00.241 Display Mode 1024 X 768 8 bits/pixel Display Refresh Rate 74 Hz (not verified) Display VRAM ? KB HD Controller RAM ? KB Hard Disk Unknown Hard Disk Controller Unknown OS Software Cache Unknown OS Software Cache Size ? KB, Windows: ? KB Off-Chip Processor Cache ? KB Operating System MS-DOS 7.00 Processor AMD-X5(tm) WriteBack Enabled, Clock Quadrupled Step 4 Features 1h Processor RAM 16 MB Processor Speed 160 MHz Sound Adapter Unknown Video For Windows Version 4.0 Windows Software Cache System Cache: write caching enabled Windows Software Cache Size All available RAM Windows System Windows 95 Enhanced Mode with Paging GX4 rev 2.0 Award 4.50G version 0306 Promise EIDE 2300 Plus controller Diamond Stealth 64 Video series 2000 (2 Meg Dram) 16 Meg Ram (70 nsec) WEIGHTED SUITE AMD 160 -2 UNITS CD-ROM WinMark 96 Not Run Thousand Bytes/Sec CPUmark16 173 [1] CPUmark32 180 [1] Disk WinMark 96 669 [1] Thousand Bytes/Sec Graphics WinMark 96 20.5 [1] Million Pixels/Sec TEST AMD 160 -2 UNITS CD-ROM/Mix Not Run Thousand Bytes/Sec Combined GWM 1 20.7 [1] Million Pixels/Sec Combined GWM 2 23 [1] Million Pixels/Sec Combined GWM 3 4.29 [1] Million Pixels/Sec Combined GWM 4 13.6 [1] Million Pixels/Sec Combined GWM 5 12.8 [1] Million Pixels/Sec Combined GWM 6 31 [1] Million Pixels/Sec Combined GWM 7 3.86 [1] Million Pixels/Sec Combined GWM 8 35.9 [1] Million Pixels/Sec Combined GWM 9 6.84 [1] Million Pixels/Sec Combined GWM10 13.1 [1] Million Pixels/Sec Combined GWM11 1.72 [1] Million Pixels/Sec Combined GWM12 0.625 [1] Million Pixels/Sec Combined GWM13 4.69 [1] Million Pixels/Sec Disk/Mix 669 [1] Thousand Bytes/Sec Processor/16-Bit 173 [1] Processor/32-Bit 180 [1] NOTES Legend: Best Score, Intermediate Score, Worst Score [1] The following Windows tasks were running during this test and could affect the test results: EXPLORER.EXE, KRNL386.EXE, MPREXE.EXE, MSGSRV32.EXE Correction to JP29 GX4 rev 2.0 Award 4.50G version 0306 Promise EIDE 2300 Plus controller Diamond Stealth 64 Video series 2000 (2 Meg Dram) 16 Meg Ram (70 nsec) The settings I came up with are based on the AMD Enhanced 120's. The cpu's pin assignments are basically the same, the only difference seems to be what the the state of the input does to the chip. (clock multiplication and cache settings). I'm confident I've enabled L1 write back cache. These settings are still preliminary as I'm still waiting for the latest bios, but at present I'm running overclocked at 160Mhz and this $70 dollar chip is outperforming the $270 Pentium Overdrive 83 Mhz chip in every category including CPU 16 and 32 mark. (using winbench 96). I'm not sure it will last at 160Mhz, but the replacement of it won't be that expensive if it does give in. I know the next one will always work at 133Mhz. If these settings can be changed, I'm welcome to suggestions as I don't claim to have all of the answers. They are proven on my system. Bios settings: Auto Config Disabled Fast Reset Emul Enabled AT Bus Clock 1/4 CLKIN(1/5 for 160 Mhz) DRAM Speed Fastest DRAM Write WS 0WS DRAM Write CAS 1T DRAM Write Burst Enabled Slow Refresh Disabled Hidden Refresh Enabled L2 Cache Scheme Write Back L1 Cache Scheme Write Back Cache Burst Read 1T Cache Write Cycle 2T Video Shadow Non-cache Memory Hole Disabled Fast Reset Latency 2 usec Latch Local Bus T2 Local Bus Ready Transparent for 133Mhz Sychronize for 160Mhz Jumper Settings: JP5 1+2 short JP6 1+2 short JP11 2+3 short JP16 1+2, 4+5 short JP17 1+2, 5+6 short JP18 1+2 short JP19 2+3, 4+5 short JP20 1+2 short for 4x clock, All open for 3X clock JP21 3+4 short (should be write back cache), Open for write thru JP22 1+2, 4+5 short JP23, JP24, JP25 33Mhz for 133 operation, 40Mhz for 160 operation JP26 2+3 short JP28 1+2 short for 133Mhz, 2+3 short for 160Mhz JP29 1+2 short for 133Mhz, 2+3 short for 160Mhz JP32 1+2 short Just let me know if this worked for you. ASUS can give me a job! difilipp@whidbey.net At 12:30 PM 3/28/96 -0800, you wrote: > >Hey John, > >I'm in the process of upgrading my BIOS and getting the 133Mhz processor >to run at 160Mhz. Here is my current thread to two other people about the >process. Once I have ALL the details, I plan on making a master GX4 >upgrade document and posting it on the ASUS newsgroup. > >Could you give the technical specs a once-over and let me know if you >dis-agree with anything? > >I greatly appreciate it and I'll talk to you soon! > > >--David > +----------------------------------------------------------------------------+ > | David A. Ranch - ISDN/Cablemodems/PCs dranch@ecst.csuchico.edu | > +---- ----+ > +----- For more detailed info, check out: www.ecst.csuchico.edu/~dranch -----+ > >---------- Forwarded message ---------- >Date: Thu, 28 Mar 1996 14:27:01 -0500 (EST) >From: Roman Shapiro >To: "David A. Ranch" >Subject: Re: Setting for AMD 5x86-133 & GX4 : I've got the processor now! > >Here's something I read in a post recently: > >++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ >Hi >I've seen numerous requests for information on how to set the jumpers >correctly for this combination. I received two of these processors this >weekend, and managed to get them going using the following jumper >settings: > >JP16: 1-2 4-5 >JP17: 1-2 5-6 >JP18: 1-2 >JP19: 1-2 3-4 5-6 >JP20: 1-2 >JP21: 2-3 >JP22: 1-2 4-5 > >If you leave JP20 open the CPU runs in clock trebling mode instead of >clock quadrupling mode. You could experiment and try to get it to work >as a 3*50MHz system. Both CPUs ran flawlessly at both 133MHz (4*33MHz) >and 160MHz (4*40MHz). The only problem I did encounter was that QEMM >7.5 hung when it started, no matter what the speed of the CPU was. >I think it has something to do with QEMM identifying the CPU incorrectly. >Cheers >Graham Inggs >+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ > > > >On Thu, 28 Mar 1996, David A. Ranch wrote: > >> >> Hey Roman, >> >> Welp, I broke down and bought a 133Mhz processor. I'm in the process of >> burning in the v401 BIOS and I would to pick your brain about what the >> jumper settings are for the GX4 motherboard. Again, thanks for the help >> and I plan on posting all my results to the ASUS newsgroup. >> >> (from you earilier) >> >> > If you have the manual, I can tell you which entry to use (like AMD >> > Enhanced). >> >> Ok, the first thing is.. this processor is called the 5x86-133. The Asus >> WWW site keeps mentioning two AMD DX4 processors: SV8B and the NV8T. >> I'm under the impression that this 133Mhz processor is neither. Can you >> clear this up for me? Ok.. for the processor stuff, check out the bottom >> of this e-mail since I'll put all the differentGx4 processor types there >> and I'll include the readme file that came with the bios upgrade just in >> case. >> >The SV8 and SV8 stuff only applied to the 80-120 Mhz processors. The 133 >has and ADZ somewhere I think. > >> > With the chip running at 4x40, I have the speed to fastest and fewest >> > wait states (I'm pretty sure). But this is something that you have to >> > tweak in the end by yourself. At 4x33, you should be able to set things >> > to max speed. >> >> Great.. I'm running an AMD Dx/2 80 5V now and I can only run it at FASTER. >> I still might not be able to get my system to run at fastest but bah.. >> thats ok. Two question(s): >> >> 1) ASUS says that you MUST run the L1 cache in >> Write-Through mode instead of the preferable Write-Back mode if >> you have a bus-mastering card like a SCSI card (which I do). >> Have you had to deal with this or it doesn't apply for your >> situation? >> >Not sure. I use write-back and have everything except for a SCSI card. > >> 2) What settings are you using for your VLBus? >> >> - VESA delay ENABLED? (ASUS default is enabled) >> - <33Mhz & 0ws (This is what I'm running now) >> or >> >33Mhz & 1ws >> >At 40Mhz, I have the VESA delay enabled. It may work without it at 33Mhz. > >> >> OK.. lastly,,from the manual as you requested, here are the different CPU >> jumper settings. Could you note to me which I need to run the processor >> at both 133Mhz and 160Mhz for our GX4 motherboards? >> > >The same setting should allow you to run at 133 and 160- you just have to >change the bus speed (from 33 to 40). Compare the settings from the >beginning of this message to the differenct cpu setting from the manual. >I think it's closest to the AMD486DXL4 - 2x clock- you may not have to >change the jumpers much. :) There may be one jumper different- the one >that enables internal write-back. > >Also, be warned that at 160Mhz (4x40), my GX4 boots up and says 150Mhz. > >Good luck. >-Roman > >> --- >> >> probably no - SL Enhanced 486SX & SX2 or Non-SL 486 & SX2 >> >> probably no - SL Enhanced (non-SL) 486DX, DX2, 487SX & ODP; 486Dx/4 @ >> 3x clock >> >> probably no - 486DX4 - 2.5x clock >> >> probably no - 486DX4 - 2x clock (BIOS reports AM486DXL4 3x) >> >> possibly - P24D Write-Back L1 cache >> JP21 Short 2&3 - Write-Through L1 cache >> >> possibly - P24CT - Write-back L1 cache >> JP21 Short 2&3 - Write-Through L1 cache >> >> possibly - P24T - Write-back L1 cache >> JP21 Short 2&3 - Write-Through L1 cache >> 237-pin SL ODP >> >> possibly 120Mhz - 486DX4 ODP (OverDrive Processor?) - 3x clock >> >> possibly - 486DX4 ODP (OverDrive Processor?) - 2.5 clock >> >> doubtful - 486DX4 ODP (OverDrive Processor?) - 2.0 clock >> >> doubtful - AMD486SXL, AMD486SX2L >> >> doubtful - AMD486DXl, AMD486DX2L >> >> possible 120Mhz - AMD486DXL4 - 3x clock >> >> This is what I - AMD486DXL4 - 2x clock >> run my Dx/2 80 >> at >> >> doubtful - Cyrix 486DX, DX2 >> >> doubtful - Cyrix 486DX2-V >> >> doubtful - Cyrix DX5 - 3x clock >> >> doubtful - Cyrix DX5 - 2x clock >> >> >> ======= >> >> >> This is the text file that came with the BIOS upgrade image. I'm >> including this just in case it helps. >> >> ---- >> >> BIOS file : SV2G0401.AWD >> BIOS Ver. : 0401 (#401A0-0401) >> Related Mainboard : VL/I-486SV2G(X4) Rev. 2.0 or above >> Related Chipset : SiS 85C471 chipset >> >> >> Description : >> >> 1. To support the following CPUs : >> a.) AMD 486DX4 (SV8B) : with green & L1 cache write-back. >> b.) Intel 486DX4 (&EW) : with green & L1 cache write-back. >> c.) Cyrix 486DX4 : with green & L1 cache write-back. >> d.) Cyrix 5x86 (M1sc) : with Pentium code inside. >> e.) AMD 486DX4 (NV8T) : non-green & L1 cache write-through. >> >> The new CPUs jumper setting as the follow : >> >> A) Cyrix 486DX4 P/O >> Cyrix 5x86 (M1sc) >> >> JP16 : 1-2,5-6 (short) >> JP17 : 1-2,5-6 (short) >> JP18 : 1-2 (short) >> JP19 : 1-2,3-4,5-6 (short) >> JP20 : (none) >> JP21 : (none) >> JP22 : 1-2 (short) >> >> >> B) AMD 486DX4 (SV8B) >> Intel 486DX4 (&EW) >> >> JP16 : 1-2,4-5 (short) >> JP17 : 1-2,5-6 (short) >> JP18 : 1-2 (short) >> JP19 : 1-2,3-4,5-6 (short) >> JP20 : (none) >> JP21 : 3-4 (short) >> JP22 : 1-2,4-5 (short) >> >> C) Cyrix 486DX4 >> >> JP16 : 2-3,5-6 (short) >> JP17 : 2-3,5-6 (short) >> JP18 : 2-3 (short) >> JP19 : 2-3 (short) >> JP20 : 2-3 (short) >> JP21 : 1-2 (short) >> JP22 : 1-2 (short) >> >> D) AMD 486DX4 (NV8T) >> AMD 486DXL4 >> >> JP16 : 1-2,5-6 (short) >> JP17 : 4-5 (short) >> JP18 : 4-5 (short) >> JP19 : (none) >> JP20 : 4-5 (short) >> JP21 : (2-3 short : 2x CLK, 2-3 open : 3x CLK) >> JP22 : 1-2, 4-5 (short) >> >> >> --David >> +----------------------------------------------------------------------------+ >> | David A. Ranch - ISDN/Cablemodems/PCs dranch@ecst.csuchico.edu | >> +---- ----+ >> +----- For more detailed info, check out: www.ecst.csuchico.edu/~dranch -----+ >> >> >> > > > John and Nellena Difilippo difilipp@whidbey.net