Major update on the Asus A8R-MVP cpu bug

jpinard · Mar 25, 2006, 03:07 AM

I'm finally at liberty to say that the big huge long threads that I started here and on Rage3d about the Asus A8R-MVP + CAT 6.2/6.3 cpu hits had been confirmed by ATI labs several weeks ago. After all that testing... and all that time... we discovered it was an Asus motherboard defect, not a problem with the Catalyst drivers. The newer Catalyst 6.2/6.3 drivers enhanced an existing problem with the Asus motherboard on X1800/X1900 boards with 512 Meg of RAM. [color=black]It is a single CPU register being incorrectly set by the ASUS bios, which results in writes across the PCIE bus going at half speed. [/color]

So I found the problem with the board/card - but ATI did the engineering to resolve the problem. Asus has sent an updated BIOS to ATI that does correct the issue, but stability tests have not been completed and there's no information on when/if this BIOS will be posted.

The irony? After the hundreds of hours I put into narrowing down this problem, the extra work ATI had to do to with their drivers to isolate the exact problem - is if Asus had simply responded to my initial requests for help I'd never have even posted these problems here. Lack of proper testing and (in my own opinion, not ATI's opinion) poor engineering has made the A8R-MVP and the A8R32 needlesssly problematic for everyone. If Asus bothered to monitor their own forums, respond to tech e-mails, read their own webforms... they could have run a simple test and seen this too. But instead it took a lot of effort from a lot of people to get Asus to listen. This is not satisfactory for a Tier-1 manufacturer, and to me, there appears to be massive disconnect between levels of managment, support and engineering at Asus.

I'd like to thank ATI, DH, and Rage3d for providing an outlet for our frustrations and concerns, without the help of which many of us Asus A8R(32)-MVP users would have been (for lack of a better term) screwed.

It's a real tribute to ATI that unlike many hardware companies, they closely monitor forums and work honestly and openly to address problems. If only all hardware venders had this kind of concern us end-users would be in a much better situation.

I think competition can sometimes bring out the best in companies and I hope DFI, Gigabyte/MSI, A-Bit have monitored the Asus situations and are preparing to capitalize on these mis-steps. I do wish there was a more constructive way to post these frustating series of events in relation to Asus - but honestly there's just no way. Everyone makes mistakes and nothing can be perfect, but it's obvious the last 6 months of Asus mobo releases does not fall in the "accidents happen" category. I'll be re-posting this note in all the relevant threads.

Crossfirepain · Mar 25, 2006, 02:14 PM

Well, I for one would be very interested in this bios...
Would rolling back to one of the older bios's fix it?

Does this bug also affect the new a8r32 board?

I would maybe be willing to test the new bios

thanks for all your work..

Crossfirepain

jpinard · Mar 25, 2006, 11:24 PM

Quote:

Originally Posted by Crossfirepain

Well, I for one would be very interested in this bios...
Would rolling back to one of the older bios's fix it?

Does this bug also affect the new a8r32 board?

I would maybe be willing to test the new bios

thanks for all your work..

Crossfirepain

I don't have it. The only people that have it are Asus and ATI. And of course ATI can't do anything with it besides see if it fixes or crashes.

Crossfirepain · Mar 25, 2006, 11:30 PM

That's why i was kinda hoping it WAS ati's driver at fault, then at least we would get a fix

Curious142 · Mar 25, 2006, 11:57 PM

It's been my experience in the last few weeks that the tech support people at both Asus and ATI are pretty much clueless. It took me three and a half weeks just to get Asus to send me an up-to-date bios chip for my P5WD2 so that it could handle an Intel dual-core cpu. But ever since I got the moboard running with the new bios chip I get a bsod when going into windows if I have an ATI driver installed. I've tried both an HIS x1800xl and a Sapphire x1800xt. ATI tech support doesn't have a clue about what's causing the problem. So I'm stuck with a machine that will only run in vga "mode" and can't handle pci-e "mode". Such a deal, a $300 vid card that runs like a $40 card. Maybe the Asus moboard needs another bios update.

CB

H82LUZ · Mar 26, 2006, 03:37 PM

Your power supply needs to be a single channel 12 v rail.The one you are using has 4,It is a kown problem with That motherboard ,Do you have a 500 watt power supply with one 12v rail and at 30 -38 amps ? It should cure your problem .It was on serval hardware sites when the X1900XT-X came out .As for the AR32 no the bug is not present on it.

http://www.theinq.net/?article=29940

and here

http://www.bit-tech.net/hardware/200...ossfire/4.html

jpinard · Mar 27, 2006, 09:29 AM

Quote:

Originally Posted by H82LUZ

Your power supply needs to be a single channel 12 v rail.The one you are using has 4,It is a kown problem with That motherboard ,Do you have a 500 watt power supply with one 12v rail and at 30 -38 amps ? It should cure your problem .It was on serval hardware sites when the X1900XT-X came out .As for the AR32 no the bug is not present on it.

http://www.theinq.net/?article=29940

and here

http://www.bit-tech.net/hardware/200...ossfire/4.html

Who are you talking to? With the issue I listed, it doesn't matter what PSU you have. You could have a billion Watt PSU, and a trillion amps on a single 12V rail. Won't matter. It' s a cpu register bug that hits older processors worse than newer ones. Plus I'm not running Crossfire, and that problem is a dual card issue... in addition, people have fixed their Crossfire issues with the FSP Group PSU. If the PSU can manage cross-load amperage fast/effectively enough... you can have a 20-rails. But no one outside FSP has managed to spilt the load across 2+ rails very well.

MindlessOath · Mar 27, 2006, 07:48 PM

MSI is not good either. they never fix problems at all. stay away from those guys.

jpinard · Mar 27, 2006, 08:08 PM

Quote:

Originally Posted by MindlessOath

MSI is not good either. they never fix problems at all. stay away from those guys.

Well that's good to know. So that leaves us with what... Gigabyte and DFI? and maybe A-Bit if they can make their boards stable again?

MindlessOath · Mar 28, 2006, 03:02 PM

i like abit boards, but never experianced the customer support. gigabyte i have always bought, i was going to return a board for repair but i never got around to actually shipping it, it went pritty smooth up to that point tho.

jpinard · Mar 29, 2006, 06:26 AM

Looks like they took the 0402 BIOS down again.

Crossfirepain · Mar 29, 2006, 07:30 AM

I don't think they take it down, it has something to do with the server the bios is on not always being available, so it sometimes doesn't show up.. I think this was discussed on asus's OWN forum for the a8r-mvp before IT dissapeared. didnt want people to keep ripping the board, I was at fry's today to pick up some memory, and i saw the a8r32-mvp, was sooooo tempted to pick it up. There is some hex editor program thing ( forget what it's called) that i used on a different MOBO. (it had a thing where hidden memory timing stuff was set as such that it was killing performance) and i was able to change the registers to correct the problem. I bet we could change the register that is off with this program, if we knew what it was... I wish they would release the bios...

Crossfirepain · Mar 30, 2006, 02:00 AM

Was it ever confirmed that all available bios's for the a8r-mvp have this bug in it?

jpinard · Apr 2, 2006, 05:50 PM

Quote:

Originally Posted by Crossfirepain

Was it ever confirmed that all available bios's for the a8r-mvp have this bug in it?

I never ran anything other than 402. It's sickening how low a priority us A8R-MVP users are to Asus.

Crossfirepain · Apr 3, 2006, 06:23 AM

Ah well, I am settings my sights on one of the new sapphire 580 based boards.

looks sweet, this is ridiculous having to buy a new mobo tho

http://www.zipzoomfly.com/jsp/Produc...uctCode=249402

jpinard · Apr 3, 2006, 06:28 AM

Quote:

Originally Posted by Crossfirepain

Ah well, I am settings my sights on one of the new sapphire 580 based boards.

looks sweet, this is ridiculous having to buy a new mobo tho

http://www.zipzoomfly.com/jsp/Produc...uctCode=249402

What Southbridge does that board use? It looks like there's a VIA chip in the screenshot.

AlexM105 · Apr 3, 2006, 07:02 AM

Quote:

Originally Posted by jpinard

What Southbridge does that board use? It looks like there's a VIA chip in the screenshot.

SB 450.

http://www.sapphiretech.com/en/produ...s.php?gpid=141

Is says it supports SATAII though, which the SB 450 does not, so it might have extra SATA ports controlled by another chip.

Mar 25, 2006, 03:07 AM	#1
jpinard DriverHeaven Addict Join Date: Jul 2003 Posts: 278 Rep Power: 0	Major update on the Asus A8R-MVP cpu bug I'm finally at liberty to say that the big huge long threads that I started here and on Rage3d about the Asus A8R-MVP + CAT 6.2/6.3 cpu hits had been confirmed by ATI labs several weeks ago. After all that testing... and all that time... we discovered it was an Asus motherboard defect, not a problem with the Catalyst drivers. The newer Catalyst 6.2/6.3 drivers enhanced an existing problem with the Asus motherboard on X1800/X1900 boards with 512 Meg of RAM. [color=black]It is a single CPU register being incorrectly set by the ASUS bios, which results in writes across the PCIE bus going at half speed. [/color] So I found the problem with the board/card - but ATI did the engineering to resolve the problem. Asus has sent an updated BIOS to ATI that does correct the issue, but stability tests have not been completed and there's no information on when/if this BIOS will be posted. The irony? After the hundreds of hours I put into narrowing down this problem, the extra work ATI had to do to with their drivers to isolate the exact problem - is if Asus had simply responded to my initial requests for help I'd never have even posted these problems here. Lack of proper testing and (in my own opinion, not ATI's opinion) poor engineering has made the A8R-MVP and the A8R32 needlesssly problematic for everyone. If Asus bothered to monitor their own forums, respond to tech e-mails, read their own webforms... they could have run a simple test and seen this too. But instead it took a lot of effort from a lot of people to get Asus to listen. This is not satisfactory for a Tier-1 manufacturer, and to me, there appears to be massive disconnect between levels of managment, support and engineering at Asus. I'd like to thank ATI, DH, and Rage3d for providing an outlet for our frustrations and concerns, without the help of which many of us Asus A8R(32)-MVP users would have been (for lack of a better term) screwed. It's a real tribute to ATI that unlike many hardware companies, they closely monitor forums and work honestly and openly to address problems. If only all hardware venders had this kind of concern us end-users would be in a much better situation. I think competition can sometimes bring out the best in companies and I hope DFI, Gigabyte/MSI, A-Bit have monitored the Asus situations and are preparing to capitalize on these mis-steps. I do wish there was a more constructive way to post these frustating series of events in relation to Asus - but honestly there's just no way. Everyone makes mistakes and nothing can be perfect, but it's obvious the last 6 months of Asus mobo releases does not fall in the "accidents happen" category. I'll be re-posting this note in all the relevant threads. __________________ [COLOR=blue]DFI LP UT NF4 Ultra-D ... AMD 64 FX-57 @2.85 Gig no extra voltage ... 2 Gig Patriot 3200LL RAM 2.5-3-2-5 1T ... ATI X1900 XTX ... Creative Labs X-Fi Xtrememusic ... Forton FX600-GLN 600 Watt[/COLOR]

Mar 26, 2006, 03:37 PM	#6
H82LUZ HardwareHeaven Lover Join Date: Jan 2004 Location: Cobourg ,Ontario ,Canada Posts: 114 Rep Power: 55 System Specs	Your power supply needs to be a single channel 12 v rail.The one you are using has 4,It is a kown problem with That motherboard ,Do you have a 500 watt power supply with one 12v rail and at 30 -38 amps ? It should cure your problem .It was on serval hardware sites when the X1900XT-X came out .As for the AR32 no the bug is not present on it. http://www.theinq.net/?article=29940 and here http://www.bit-tech.net/hardware/200...ossfire/4.html __________________ Last edited by H82LUZ; Mar 26, 2006 at 06:15 PM.

Mar 27, 2006, 07:48 PM	#8
MindlessOath DriverHeaven Senior Member Join Date: Jul 2003 Posts: 1,121 Rep Power: 0	MSI is not good either. they never fix problems at all. stay away from those guys. __________________ ATI Radeon X800XT PE <---- LOOK UPDATED!! AMD Athlon 2Ghz, nforce2, 512mb ddr3500 LINKS U SHOULD KNOW!!!! DriverCleaner - use every time u install new drivers! ATI driver feedback - to report driver problems Tacticalcenter - my Gaming website and forums ATI 3rd party Optimized Driver Compairison

Mar 28, 2006, 03:02 PM	#10
MindlessOath DriverHeaven Senior Member Join Date: Jul 2003 Posts: 1,121 Rep Power: 0	i like abit boards, but never experianced the customer support. gigabyte i have always bought, i was going to return a board for repair but i never got around to actually shipping it, it went pritty smooth up to that point tho. __________________ ATI Radeon X800XT PE <---- LOOK UPDATED!! AMD Athlon 2Ghz, nforce2, 512mb ddr3500 LINKS U SHOULD KNOW!!!! DriverCleaner - use every time u install new drivers! ATI driver feedback - to report driver problems Tacticalcenter - my Gaming website and forums ATI 3rd party Optimized Driver Compairison

Mar 29, 2006, 06:26 AM	Thread Starter #11
jpinard DriverHeaven Addict Join Date: Jul 2003 Posts: 278 Rep Power: 0	Looks like they took the 0402 BIOS down again. __________________ [COLOR=blue]DFI LP UT NF4 Ultra-D ... AMD 64 FX-57 @2.85 Gig no extra voltage ... 2 Gig Patriot 3200LL RAM 2.5-3-2-5 1T ... ATI X1900 XTX ... Creative Labs X-Fi Xtrememusic ... Forton FX600-GLN 600 Watt[/COLOR]

Mar 25, 2006, 02:14 PM	#2
Crossfirepain DriverHeaven Newbie Join Date: Mar 2006 Posts: 14 Rep Power: 0	Well, I for one would be very interested in this bios... Would rolling back to one of the older bios's fix it? Does this bug also affect the new a8r32 board? I would maybe be willing to test the new bios thanks for all your work.. Crossfirepain

Mar 25, 2006, 11:30 PM	#4
Crossfirepain DriverHeaven Newbie Join Date: Mar 2006 Posts: 14 Rep Power: 0	That's why i was kinda hoping it WAS ati's driver at fault, then at least we would get a fix

Mar 25, 2006, 11:57 PM	#5
Curious142 DriverHeaven Newbie Join Date: May 2003 Location: Colorado Posts: 7 Rep Power: 0	It's been my experience in the last few weeks that the tech support people at both Asus and ATI are pretty much clueless. It took me three and a half weeks just to get Asus to send me an up-to-date bios chip for my P5WD2 so that it could handle an Intel dual-core cpu. But ever since I got the moboard running with the new bios chip I get a bsod when going into windows if I have an ATI driver installed. I've tried both an HIS x1800xl and a Sapphire x1800xt. ATI tech support doesn't have a clue about what's causing the problem. So I'm stuck with a machine that will only run in vga "mode" and can't handle pci-e "mode". Such a deal, a $300 vid card that runs like a $40 card. Maybe the Asus moboard needs another bios update. CB

Mar 29, 2006, 07:30 AM	#12
Crossfirepain DriverHeaven Newbie Join Date: Mar 2006 Posts: 14 Rep Power: 0	I don't think they take it down, it has something to do with the server the bios is on not always being available, so it sometimes doesn't show up.. I think this was discussed on asus's OWN forum for the a8r-mvp before IT dissapeared. didnt want people to keep ripping the board, I was at fry's today to pick up some memory, and i saw the a8r32-mvp, was sooooo tempted to pick it up. There is some hex editor program thing ( forget what it's called) that i used on a different MOBO. (it had a thing where hidden memory timing stuff was set as such that it was killing performance) and i was able to change the registers to correct the problem. I bet we could change the register that is off with this program, if we knew what it was... I wish they would release the bios...

Mar 30, 2006, 02:00 AM	#13
Crossfirepain DriverHeaven Newbie Join Date: Mar 2006 Posts: 14 Rep Power: 0	Was it ever confirmed that all available bios's for the a8r-mvp have this bug in it?

Apr 3, 2006, 06:23 AM	#15
Crossfirepain DriverHeaven Newbie Join Date: Mar 2006 Posts: 14 Rep Power: 0	Ah well, I am settings my sights on one of the new sapphire 580 based boards. looks sweet, this is ridiculous having to buy a new mobo tho http://www.zipzoomfly.com/jsp/Produc...uctCode=249402