Asus A8V Deluxe Technical Discussion Thread

[tutu]

Thanks for the setup help - Audigy2ZS in PCI-4, and I am going to set AI-overclocking to manual!

I am not sure how they can support the PCI/AGP lock in future BIOS revisions if it only works on some revisions of the board.

Perhaps this shows just how little testing they do before shipping a product.

[jlangner]

Does anyone play Battlefield Vietnam. I got my A8V and A64-3500 up and running last night and the game would freeze randomly. I read where some thought it was an issue with NForce boards and Punkbuster but it happened on my VIA board last night.

[DeathEater]

My system is still serioulsy unstable in BFV and simply using windows.

I have tried lots of stuff including changing voltages etc but its still playing up. May have to get the board sent back.

Last reboot windows told me the ATI graphics driver caused it. Im using Cat 4.6 - should I try an older driver?

[jlangner]

Noticed their is a new Beta Bios for the A8V - posted yesterday. A8V Deluxe Beta BIOS 1006.004.

[DeathEater]

Yes, using that bios at the moment - have tried them all with no luck on the stability front for me

[jlangner]

Check this thread. I think it is Punkbuster.

http://forums.pcper.com/showthread.p...5&page=2&pp=15

[Senor Panadero]

Just a stab in the dark,
"Instability" can mean anything, but those of you with issues might try raising the tRCD one notch from where it is now. Was a common issue in the K8V threads, especially with lower latency memory. Mushkin explains it well in their A64 report.

[jlangner]

My instability is the screen freezing every so often. Other than the A8V and A64-3500 everything else is exactly the same as it was 2 days ago with a A7N8X and XP2800.

[Arvig]

Thanks for the setup help - Audigy2ZS in PCI-4, and I am going to set AI-overclocking to manual!

I am not sure how they can support the PCI/AGP lock in future BIOS revisions if it only works on some revisions of the board.

Perhaps this shows just how little testing they do before shipping a product.

Sorry I didn't look at the PCI slot question earlier, but I got the impression that PCI slot 5, the "southernmost" one is at least partially dedicated to the Wi-Fi card. I think one can use the Wi-Fi card in any PCI slot per-se (and I'd assume any board, even a non-ASUS one), but the idea is that slot 5 has the most consistant voltage for the card. I could be wrong on this, but I think I remember reading this.

[stevenmh]

I had some trouble trying to get stable at 247x10.5. I could pass memtest/P95, but 3DMark would either crash to desktop or freeze up. I bumped V-Link from 2.5 to 2.6v and AGP from 1.5 to 1.6v, and now it's fine. At least once I got an error message regarding the video driver having failed to complete an operation or something along those lines. I don't know why either of those voltages would make a difference, since I've locked the PCI/AGP bus at 33/66 and haven't overclocked the video card (yet), but I'm just glad it's stable now. This is with Cat 4.6 and the VIA drivers on the CD included with the board.

At stock speeds, though, it was stable at the stock voltages. I've learned from experience to verify P95/3DMark stability before I start OC'ing anything, so I can go ahead and troubleshoot and/or get the RMA out of the way before I waste my time trying to find a stable OC.

[Arvig]

Just a stab in the dark,
"Instability" can mean anything, but those of you with issues might try raising the tRCD one notch from where it is now. Was a common issue in the K8V threads, especially with lower latency memory. Mushkin explains it well in their A64 report.

This might make sense, the board sees my Corsair PC3200 XMS XL Pro as 2-3-2-5, even though the memory is rated from the factory as 2-2-2-5. Using the original first release BIOS 1003, Memtest 86+ had the odd habit of resetting things to 2-2-2-5, which even at pure stock speeds, I'd occasionally bomb out, and couldn't pass the Sandra 2004 Memory bandwidth test. Any BIOS newer then 1003 doesn't do this with Memtest 86+ it seems, even with the memory on Auto.

I guess my point is that you may have to tweak the settings, and/or go to a newer BIOS...I also found I can run a 222 FSB with BIOS 1006.004 at 2-3-2-5, I had to go to slower timings (either 2-3-3-6 or 2-3-2-10) to run a 222 FSB with an older BIOS, even 1005.021. Ironically, if 1006.004 had working locks, I almost wonder if I could run 11 * 240 or 11 * 241, I couldn't without failing Memtest 86+ using my memory even though that was an acceptable OC for my CPU...seems like BIOS 1006.004 is allowing better memory timings...but of course has no way of setting the lock... oh well.

Oh, and yes, no data corruption at a 222 FSB. I'm running two SATA drives, but not a RAID array, both on the VIA ports.

Oh, since someone asked for 1005.021, here's a link. You'll probably have to join the forum this links to in order to pull down these BIOS. Go to the "ASUS BIOS for AMD" thread, then post 4. (If you see a post for a A7N8X BIOS, you've gone too far. ).

http://www.bleedinedge.com/forum/showthread.php?t=2224

Someone has already linked to 1006.004 here, and you can get the standard 1006 from ASUS.

[Arvig]

I had some trouble trying to get stable at 247x10.5. I could pass memtest/P95, but 3DMark would either crash to desktop or freeze up. I bumped V-Link from 2.5 to 2.6v and AGP from 1.5 to 1.6v, and now it's fine. At least once I got an error message regarding the video driver having failed to complete an operation or something along those lines. I don't know why either of those voltages would make a difference, since I've locked the PCI/AGP bus at 33/66 and haven't overclocked the video card (yet), but I'm just glad it's stable now. This is with Cat 4.6 and the VIA drivers on the CD included with the board.

At stock speeds, though, it was stable at the stock voltages. I've learned from experience to verify P95/3DMark stability before I start OC'ing anything, so I can go ahead and troubleshoot and/or get the RMA out of the way before I waste my time trying to find a stable OC.

What Game/Test are you bombing out on 3dMark 03, if I may ask?

Oddly enough, I just crashed and had to reset the CMOS if I bumped my AGP voltage from 1.5 to 1.6, but that was back on BIOS 1003.

Sounds like from an earlier thread we're getting about the same OC in mhz on our CPU's, but you can hit a larger memory FSB, but you also have different memory. Just kind of keeping recond in my own mind to see where people are heading, to see if my results are typical. To give an idea, I can run at 222 * 12 all day long, lock or no lock. 223 * 12, with or without lock I'd pass every test under the sun, but probably crash after an hour or two, especially if running a game. I get SATA data corruption bad enough to sometimes force me to do a Windows reinstall at 224 FSB, unless I use a BIOS that allows me to set the lock. Oddly enough, my memory bandwidth scores at 222 * 12 and 232 * 11.5 are the same, and Memtest 86+ actually claims my memory is faster at 222 * 12 compared to 232 * 11.5. As stated earlier, even if I pump up voltage and run slow timings, I can't quite do 240 * 11, I fail memtest 86+.

[Arvig]

My system is still serioulsy unstable in BFV and simply using windows.

I have tried lots of stuff including changing voltages etc but its still playing up. May have to get the board sent back.

Last reboot windows told me the ATI graphics driver caused it. Im using Cat 4.6 - should I try an older driver?

My dad has this board too and has no problems with Cat 4.6, but I did stick with Cat 4.5 since 4.6 was showing lower benchmarks in everything but 3dMark03...for some reason, Cat 4.6 really improves benchmark scores in 3dMark03, but makes everything else I have score poorly.

[*Fab_DPRGI*]

hi guys, i've been reading the post since its birth, i too have a a8v and 2x512 corsair xl 3200 pro.

i have a mb450b mobo, so the 1005.021 doesn't work (i tried anyway...)

but i received this (quite) confortant reply from asus technical support

-------------------
Hello,

I just bought a A8V deluxe 939 wi-fi Asus motherboard, with a k8 3800+
prococessor.
I must complain about the "pci/agp fix" setting. It simply doesn''t work,
when I try to set it to 33/66, the system doesn''t boot, even with all
other
parameters set on default value.

I bought this board (quite) only for this option, as the manual clearly
states its presence.

I tried every bios(release &betas), without success.

Board serial is 45MG0B7757. I saw in some forums some disturbing info
about
all boards prior to 45MG0Jxxxx not working for this option.

If true, I don''t mind whose is the fault, I simply want my board working
with this feauture.

Can you confirm such rumors ? If so, I think to deserve a working board,
so
could you please indicate the correct procedure for RMA ? I bought it from
the italian shops ESSEDI in Milan.

If you don''t confirm, I could only take I t back to the reseller as "not
working", repeat the process until a working one is found. but, to be
honest, I am sure that a brand lie asus will not leave me in the dust, and
will prompt me with a seriuos solution to this problem.

Please don''t respond evasivly, take your time if needed. Thanks.

Fabrizio Botti, Asus consumer.



Email solution
-----------
This is a problem that we are currently addressing on this board. It
should be corrected and fully operational on the next official BIOS
revision. The revision will be 1007 and should be out within the next few
weeks. Thanks.

----------------------------

with the latest 1006 official bios, i reached 223 fsb (haven't tried yet more, will do next days) at timings 2-3-5-2-1t. my proc is a 3500+, so at 223*11 i get (windows xp) :

sandra mem : 6500/6392
super pi : 35 secs.

[stevenmh]

What Game/Test are you bombing out on 3dMark 03, if I may ask?

Oddly enough, I just crashed and had to reset the CMOS if I bumped my AGP voltage from 1.5 to 1.6, but that was back on BIOS 1003.

Sounds like from an earlier thread we're getting about the same OC in mhz on our CPU's, but you can hit a larger memory FSB, but you also have different memory. Just kind of keeping recond in my own mind to see where people are heading, to see if my results are typical. To give an idea, I can run at 222 * 12 all day long, lock or no lock. 223 * 12, with or without lock I'd pass every test under the sun, but probably crash after an hour or two, especially if running a game. I get SATA data corruption bad enough to sometimes force me to do a Windows reinstall at 224 FSB, unless I use a BIOS that allows me to set the lock. Oddly enough, my memory bandwidth scores at 222 * 12 and 232 * 11.5 are the same, and Memtest 86+ actually claims my memory is faster at 222 * 12 compared to 232 * 11.5. As stated earlier, even if I pump up voltage and run slow timings, I can't quite do 240 * 11, I fail memtest 86+.

There wasn't any particular game or benchmark. Sometimes it would make it over halfway through and freeze, sometimes it would freeze 10 seconds after firing it up.

I've pushed my system too far several times, but I've never had to clear the CMOS. The recovery feature has worked very well for me...when I go too far all I have to do is cut off the power supply, turn it back on, and then it'll start up with the message that it failed due to overclocking and let me go into the BIOS and make changes.