A8N-SLI "best" BIOS

gobnu1

Weaksauce
Joined
Jan 12, 2003
Messages
94
Having NO luck overclccking my board with the 1002 BIOS. Which of the beta BIOS seems to offer the best OC chance with stability? Also is there a reason why the 1002 BIOS won't OC at all or is there a secret to OC'ing this board?
 
My board shipped with the 1002 bios. I was able to oc using those bios but still switched to the 1003.005 beta bios. I am also able to oc with these. Sounds like a setting in bios. Have you changed your HT setting to 3 instead of 5. I've also noticed this board is very responsive to ram timing settings. Might want to go over those again and loosen things up.
 
tried a basic OC, left memory in auto changed HTT to 4x and upped cpu from 200 to 220. choked at BIOS screen and when I lowered it to stock HD was corrupt. Also clear CMOS jumper didn't work had to take out battery to clear CMOS.
 
Mine choked until I uped the voltage. Now it is running at 250 without a problem. I flashed to 1003.5 and it seems to work fine. They added a PCI Lock setting. Not sure if it worked in 1002. I did have to reformat my SATA drive twice from curroption with the 1002 bios.
 
The 1002 BIOS is a piece of shit, you cant overclock with it as the PCI-E lock does not work. I am using the 1003 005 BIOS at present I am at 2400Mhz on 1.5v.

Has anyone heard about Prime95 issues with the Winchester chips?

Also have you guys checked that your PCI cards are running at 16x, because mine and several other are running at 8x. I think they have the SLI selector in the dual card position!
 
Mine PCI-E socket came out of the box setup for dual had to flip card over. Thanks for all the replies so far will try tonite and let you all know.
 
GooN3r said:
The 1002 BIOS is a piece of shit, you cant overclock with it as the PCI-E lock does not work. I am using the 1003 005 BIOS at present I am at 2400Mhz on 1.5v.

Has anyone heard about Prime95 issues with the Winchester chips?

Also have you guys checked that your PCI cards are running at 16x, because mine and several other are running at 8x. I think they have the SLI selector in the dual card position!

I tried the 1003.006 BIOS linked above, and it did the same thing the other 1003.00x BIOSes did, and that's not see my SATA drives connected to the NVidia chipset ports. It did see them for a bit, not drop them immediatly, so that tells me that it's possibly a known issue, but still...

For the record, I'm using two Maxtor Diamond Max 10 250gb drives with NCQ. I think NCQ is the issue, I saw that someone else using drives with NCQ, specfically a Maxtor Diamond Max 10 had the NVidia ports not be able to see the hard drive with any of the 1003.00x BIOSes.

As for overclocking, I can set my FSB up to 210 and not have the BIOS tell me "System Fail due to CPU Overclocking" via the speach post reporter. Any higher FSB, regardless of voltage and it acts like I'm at too high of an overclock. If I set it to halt on no errors, I can force it to accept a higher one, but I have to manually enter it everytime in the BIOS, else it posts at stock speeds. On the 1003.00X BIOSes I could set a higher FSB and not have it halt, no problem...but of course, that does me no good if it won't see my hard drives.

Haven't tried the Silicon Image SATA ports, but since I don't want to reformat or experiment with doing something weird like putting the drivers on the drives then trying to boot up (especially since I'm using a RAID 0 with the NVidia ports right now), I doubt I will. Guess I'll wait for 1004 or whatever the next "final" BIOS will be, and hope my drives will be recognized. Maybe I'll shoot off a letter to ASUS for comedy relief, explaining the situation, and see what sort of reply in "Engrish" I get, and see if the guy even addresses this issue.

(not flaming ASUS per-se, just seems like when I send off a letter to any mobo maker, regardless of brand, I get something back that has nothing to do with anything, and/or is a bunch of fluff that is designed to placate without answering anything)

Me
 
I had no problem with the 1003 005 BIOS seeing my SATA raptors in a RAID 0 until I switched my the SLI selector from dual card to single card. The card now runs in 16x but I sometimes get boot problems caused my my hard disk drive :mad:

Guess that the price we pay, for fook sake Nvidia pull your finger out!
 
If you use nvidia sata ports, watch out for data corruption, that said...
1002 bios WILL overclock.
Set DDR timings to auto but set the speed to DDR400 manually, this will give really loose timings but allow to overclock one to one REAL HIGH.
Set the HT to 4x.
Set the pci-e to 101mhz will turn on the pci lock.
Vcore is funky; Setting vcore to 1.43 will get vcore of 1.53. A friend that has one tryed 1.6 it returned 1.43. *1.41 = 1.41 which got my 3000+ up to 260fsb stable, but I got sata corruption. Could be my fault, weak powersupply.
Be sure to set everything manually in the jumper free area, or weird things will happen.

Side note, with 1002 bios, seems to be another bug... Every boot will say overclocking falure and if you turn off halting on errors, it will boot at stock settings. On every boot just have to go into bios, and exit with or without saving will cause it to boot correctly overclocked.
 
vxspiritxv said:
Side note, with 1002 bios, seems to be another bug... Every boot will say overclocking falure and if you turn off halting on errors, it will boot at stock settings. On every boot just have to go into bios, and exit with or without saving will cause it to boot correctly overclocked.


I probably described it poorly, but that's what I got, at least if I go over a FSB of 210. 210 and under, it's actually fine.
 
There are no good BIOS available at the mo, its a case of choosing the best of a bad bunch, come Nvidia sort it out.
 
installed 1003-005 BIOS and am curently at MEM auto CPU 250X10 HTT 3 and currently benchmarking. Update in the am
 
gobnu1 said:
installed 1003-005 BIOS and am curently at MEM auto CPU 250X10 HTT 3 and currently benchmarking. Update in the am

Bet you dont have any SATA drives!!!

Right back to the stable un-overclockable 1002 official BIOS, guess I will have to wait until a decent official release arrives.

Reading the forums all the beta BIOS seem to have problems detecting SATA drive which does not help me with 2xWD Raptors
 
You lose the bet, I have a 250gig SATA Western Digital drive. Still playing with settings trying to find fastest combo.
 
No RAID 0 then, its was one or the other. IMO the beta's are just too unstable gonna wait for an official release.

At least I know my chip will do 2400Mhz easily!
 
GooN3r said:
Bet you dont have any SATA drives!!!

Right back to the stable un-overclockable 1002 official BIOS, guess I will have to wait until a decent official release arrives.

Reading the forums all the beta BIOS seem to have problems detecting SATA drive which does not help me with 2xWD Raptors

I have no idea why you are claiming 1002 isn't overclockable... I run my 3000+ at 2.55ghz 24/7 and I can get suicide screenshots of 2.75ghz (at this point I am 100% certain it is my cpu which can go no further)

Anyway I HIGHLY recommend using clockgen to overclock. As long as you set the FSB to 201 and the PCI-E bus to 101 in the BIOS it locks everything (now I am pretty sure all you need to do is set the FSB to 201, but I do it to both because I feel safer)

Nforce4 Clockgen:
http://www.cpuid.com/download/CG-NVNF4.zip

Also, DISABLE everything you don't use.

I am using 1002 because every Beta Bios I have tried fails to recoginize my DiamondMax 10 drive... The only thing the Beta Bioses have going for them is they make it a little easier to overclock, but if you know what you are doing, you can get the same clock speeds in 1002.

Also, the Beta Bioses DO NOT have any problems detecting my other SATA devices (I have a convereted 250gb That has a IDE to SATA adpater, and two 120gb WD SATA drives)... It seems it is only the DiamondMax 10's or sometimes I have heard just standard old school Maxtor SATA drives are failed to detect... Plenty of people run raptors on the Beta Bioses...
 
ok I am sticking with the 1002 bios right now, I need some help getting better memory speed as I am running my 3200+ at 10x260 and my ram is showing in cpuz as 133???
 
J-Mag said:
I have no idea why you are claiming 1002 isn't overclockable... I run my 3000+ at 2.55ghz 24/7 and I can get suicide screenshots of 2.75ghz (at this point I am 100% certain it is my cpu which can go no further)

Anyway I HIGHLY recommend using clockgen to overclock. As long as you set the FSB to 201 and the PCI-E bus to 101 in the BIOS it locks everything (now I am pretty sure all you need to do is set the FSB to 201, but I do it to both because I feel safer)

Nforce4 Clockgen:
http://www.cpuid.com/download/CG-NVNF4.zip

Also, DISABLE everything you don't use.

I am using 1002 because every Beta Bios I have tried fails to recoginize my DiamondMax 10 drive... The only thing the Beta Bioses have going for them is they make it a little easier to overclock, but if you know what you are doing, you can get the same clock speeds in 1002.

Also, the Beta Bioses DO NOT have any problems detecting my other SATA devices (I have a convereted 250gb That has a IDE to SATA adpater, and two 120gb WD SATA drives)... It seems it is only the DiamondMax 10's or sometimes I have heard just standard old school Maxtor SATA drives are failed to detect... Plenty of people run raptors on the Beta Bioses...

Can confirm, as I've stated above and in other threads that Diamond Max 10's have problems with the beta BIOSes. I think it's NCQ drives in general.

Even if you don't want to use Clockgen, I can get to a 210 FSB before it complains and tells me "system fail due to CPU overclocking", so one can do a mild overclock at least even in the BIOS.
 
Using the 1002 I am able to oc but upon reboot I have to enter bios and back out to boot. I am still using the 1003.005 bios. Using the Manual setting I am able to go fsb all the way up to 300 with the help of the dividers and the multi settings. I'm reaching 250x12 with 400 1T 2.5 -4-3-7 . After two days of no "new" problems I had to change power supplies and also chang bios back to 1002. Not sure which did the trick but I'm now using the 1003.005 like I said. Going to try the 1003.006 to see what happens. Haven't had a problem with it finding my SATA drives. Using 2x74gig Raptors on the nvidia Raid0.
Vapo LS cooling.
 
Shogan191 said:
Using the 1002 I am able to oc but upon reboot I have to enter bios and back out to boot. I am still using the 1003.005 bios. Using the Manual setting I am able to go fsb all the way up to 300 with the help of the dividers and the multi settings. I'm reaching 250x12 with 400 1T 2.5 -4-3-7 . After two days of no "new" problems I had to change power supplies and also chang bios back to 1002. Not sure which did the trick but I'm now using the 1003.005 like I said. Going to try the 1003.006 to see what happens. Haven't had a problem with it finding my SATA drives. Using 2x74gig Raptors on the nvidia Raid0.
Vapo LS cooling.


Yeah, this board seems to like Raptors in general, I think it's confirmed that the beta BIOSes are fine as long as one isn't using NCQ based hard drives.
 
I've tried the retail and beta bioses but was having problems with 103.005. It would not even post half the time. Went back to 1002 and its running quite stable.
 
Stable......buty hows your overclock? I can go real high, but i will get error messeges and it tells me to go to the bios and load default. So i do, but i dont change ANYTHING, save and exit, and im fine. I can play games, whatever. And thats at 250 fsb.(with correct mem timings and voltages, etc..) Everything is great untill i power down. Then i have to repeat the process all over again. And how bout some WHQL Nforce4 chipset drivers. Lets go already, i cant take it much longer.
 
Arvig said:
Can confirm, as I've stated above and in other threads that Diamond Max 10's have problems with the beta BIOSes. I think it's NCQ drives in general.

Even if you don't want to use Clockgen, I can get to a 210 FSB before it complains and tells me "system fail due to CPU overclocking", so one can do a mild overclock at least even in the BIOS.

Why would you not want to use clockgen? It is by far superior to using the BIOS because:

A: You don't have to freaking reboot all the time.
2: You can actually up the VID therefore getting more CPU voltage than if you just used the BIOS
d: Different apps have different overclocking tolerances, so you can adjust dynamically (for instance, I can run CoD all day at 2.6ghz but Doom3 will crash at that speed if I run like 4 to 5 consecutive timedemos)
 
I have a DiamondMax10 300gb. I cannot OC much past a 210 FSB without getting immediate BSOD errors with RAM dumps. I think this may be due to the Maxtor drive being very sensitive to OCs and that the SATA is being overclocked too even though PCI-E and PCI locks are working.
 
J-Mag said:
Why would you not want to use clockgen? It is by far superior to using the BIOS because:

A: You don't have to freaking reboot all the time.
2: You can actually up the VID therefore getting more CPU voltage than if you just used the BIOS
d: Different apps have different overclocking tolerances, so you can adjust dynamically (for instance, I can run CoD all day at 2.6ghz but Doom3 will crash at that speed if I run like 4 to 5 consecutive timedemos)


Well, since you asked....

1) Don't know if it happened to you, but when I set the voltage to 1.5 with Clockgen, and then check the voltage with CPU-Z, a program made by the same people my voltage was up around 1.65 or so, so I didn't like how it set the voltage.

2) Maybe I don't have to reboot, but unless I'm missing something, there's no way to set the overclock permamently so thus every time I do restart I have to reset the OC, with a BIOS setting it's set once and if stable, I'm done.

3) I don't mind finding the one OC that works for everything, even if one app can do a couple more FSB then the other, I'd rather find one general stable one.

As an aside, my CPU is an air cooled FX-55, so it's not like it's a huge overclocking scenario anyway, not like I'm OCing a 3500+. On my previous board, I was maybe at 216 * 13 aircooled with the same CPU, so being "stuck" at 210 FSB with a FX-55 if I don't lower the multiplier isn't the end of the world for me, assuming the next final BIOS does let me use NCQ drives.
 
BigGreenMat said:
I have a DiamondMax10 300gb. I cannot OC much past a 210 FSB without getting immediate BSOD errors with RAM dumps. I think this may be due to the Maxtor drive being very sensitive to OCs and that the SATA is being overclocked too even though PCI-E and PCI locks are working.


What BIOS are you using? If I didn't mind in 1002 the fake halt on an OC error, I could OC past 210 using 2 x Diamond Max 10 250gb's, they have the same buffer size as the 300gb.
 
1003.006 beta seems to be very stable for me so far. Much better than the 1002. I'd give it a shot.
 
Dr. X said:
1003.006 beta seems to be very stable for me so far. Much better than the 1002. I'd give it a shot.

Agreed, also heard about 1003.008, which can be found on a thread on this board.

One warning, if any of your SATA drives use NCQ (Native Command Queuing), then stick with 1002 until probably the next "final" BIOS, and even then read posts on some of the hardware site boards to see what people say about it. None of the 1003.00x series BIOSes, at least in my personal experience work with NCQ drives. For the record, I have 2x 250gb Maxtor Diamond Max 10's w/NCQ & 16mb buffers, seen same complaints regarding other Maxtor drives with NCQ, and one or two with people using Seagates with NCQ also.
 
Back
Top