• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

HUGE leap forward with A8N SLI Deluxe overclocking

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Yeah, with your memory you should be able to run up to 240mhz 1T no problem. As others have reported this board doesn't seem to like ram over 240 with 1T. I can get mine to run over 240 1T, but I have to clock the CPU much lower to do it. Hopefully a future bios will fix this issue, as the OCZ you and I have has no problem running 270+ on other boards with timings like 3-3-3-10 or even 2.5-3-3-10.

Let me know how it goes with the FX55 overclock, one day I may want to upgrade to that bad boy.
 
/Begin Hijack

While we're on the subject of overclocking, perhaps you gents could provide some input:

I've gotten my A8N-SLI Deluxe rig up and going, and now I'm fiddling with the OC. System follows:

ASUS A8N-SLI Deluxe (1002)
Athlon 64 3000+ Winchester (DH8-D0) (2196Mhz, 244FSB, 4xHTT, 1.5125v)
1GB Mushkin Basic Green PC3200 (~400Mhz, 2.5-3-3-8, 2.7v, 5:4 divider)


Now, all this checks out in windows...The system Primed for 8 hours last night with no errors, warnings. It swallowed 4 hours of Half-Life 2 without breaking a sweat. When I checked the temp this morning before stopping Prime, it was at 44C, so I'm assuming that's the highest load temp it's going to hit.

Weird thing is, sometimes it boots perfectly, and sometimes it gives me a system boot fail. I've fiddled with the RAM voltage, and that doesn't seem to be the issue, so now I'm edging up CPU voltage. I imagine it's one of these two culprits - one of them simply needs more juice - but any other ideas?

I'm also considering taking the RAM back to 1:1 and seeing how high I can ride it, but I'll have to check the net a bit and see what this stuff is supposed to hit before I try that - get a ballpark idea.

/End Hijack
 
Originally posted by: Insomniak
/Begin Hijack

While we're on the subject of overclocking, perhaps you gents could provide some input:

I've gotten my A8N-SLI Deluxe rig up and going, and now I'm fiddling with the OC. System follows:

ASUS A8N-SLI Deluxe (1002)
Athlon 64 3000+ Winchester (DH8-D0) (2196Mhz, 244FSB, 4xHTT, 1.5125v)
1GB Mushkin Basic Green PC3200 (~400Mhz, 2.5-3-3-8, 2.7v, 5:4 divider)


Now, all this checks out in windows...The system Primed for 8 hours last night with no errors, warnings. It swallowed 4 hours of Half-Life 2 without breaking a sweat. When I checked the temp this morning before stopping Prime, it was at 44C, so I'm assuming that's the highest load temp it's going to hit.

Weird thing is, sometimes it boots perfectly, and sometimes it gives me a system boot fail. I've fiddled with the RAM voltage, and that doesn't seem to be the issue, so now I'm edging up CPU voltage. I imagine it's one of these two culprits - one of them simply needs more juice - but any other ideas?

I'm also considering taking the RAM back to 1:1 and seeing how high I can ride it, but I'll have to check the net a bit and see what this stuff is supposed to hit before I try that - get a ballpark idea.

/End Hijack

I have the exact same thing happen. In Windows I'm Prime95 stable at 2.55 GHz at 1.52V for 2+ hours, but when I try to boot even with a very small overclock I often get the boot error on POST.

Also, I can't seem to get Vcore much over 1.5V. The voltage settings in BIOS and in AI Booster (which actually seems to drop the voltage) aren't doing what they're supposed to. Clockgen is the only way I can get Vcore over 1.5V.

I think the board is fine, but the BIOS leaves a lot to be desired.

 
Might just be BIOS issues. I'm f*ckin with it - more later.

edit: Yeah I think it'sa BIOS thing, because it happens to me every other boot, almost regularly. I know it's not the components - This CPU primed for over 8 hours twice at this clockspeed, and it's rocksteady.

Today I ran a 2.5 hour memtest with 800% coverage, 0 errors. Neither the processor nor the memory are unstable.

As far as I can tell, it's just this BIOS. It needs to be ironed out - once it is, this board will be top grade. It's still rather good as is, all the problems are on the software end, not the hardware end.
 
What bios are you guys running. I had similar problems prior to going to the 1003.005 bios, this seems to have fixed alot of the issues, I don't get any post errors any more. Although AI booster is still messed up, and I still use Clockgen to increase voltage. At least now when I set voltage with Clockgen and reboot, it keeps my voltage changes.
 
Originally posted by: bap4201
there is a link i posted for 1003-006 beta from germany site. Have you tried/r u using this?

No I've downloaded it, but haven't flashed it yet. I noticed its a zip file, do you just unzip, then flash the resulting file?
 
I'm onm the 1002 BIOS and I'd rather not flash until ASUS releases another official BIOS that actually fixes issues, instead of these betas that may or may not do something.

I'll try clockgen and see what happens.
 
I know, at least on the 754 chipsets with my particular OCZ RAM, that the timings set by SPD are set far too aggressively for the platform to be stable at their stated speeds, i.e. DDR500. I'm not just talking about the 2.5-3-3-7 timings either. What helped me to get to 1T, and increased stability, was manually setting these timings. I'll try and look up which one's I'm talking about when I get home.
 
Ok, the saga continues and the plot thickens...

Installed ClockGen and tried it out, and things got odd. In the BIOS I have CPU voltage at 1.5v, and dimm voltage at 2.6v

CPU-Z and ASUS Probe agree with the BIOS. The CPU is getting 1.504 - 1.52v they say. ClockGen, on the other hand, says it's only pulling in 1.375v

The odd part about this is that ClockGen and CPU-Z both use the same hardware detection engine - Panopsys - so who's lying? I'm inclined to think ClockGen is the liar in this case, as I have three other utilities agreeing, plus 1.375 seems like it would be far too low for the CPU to be stable at a 400Mhz overclock.

Thoughts?
 
Yeah, Clockgen voltage readings are messed up. However you can get the highest available voltages by setting them via Clcokgen. Go figure? And if you use bios to set voltage, you will notice the AI and CPU-Z show much lower than set in bios.
 
As continued from my earlier post, the settings that seemed to help me the most, other than the standard latencies (which are 2.5-3-3-10), were the Row Cycle Time and Row Refresh Cycle Time. Set those to 15 and 17 respectively, and the RAM should be significantly more stable at higher clocks. That's how I got my memory to do 240MHz(480DDR), 1T. You can work your way down on those settings until you find the "sweet spot". For me, the fastest I got those settings to was 9 and 13 respectively.

Just thought I'd throw that out there for the people with memory issues. Hopefully it'll help you out.
 
Originally posted by: GuitarDaddy
Yeah, Clockgen voltage readings are messed up. However you can get the highest available voltages by setting them via Clcokgen. Go figure? And if you use bios to set voltage, you will notice the AI and CPU-Z show much lower than set in bios.



Actually, my ASUS Probe and CPU-Z agree with the BIOS - BIOS is at 1.5v, and that's what they both report, although it bounces around between 1.488 and 1.520, a little more fluctuation than I'd like, but within acceptable limits considering stock voltage on this chip is like 1.4v - it's getting plenty of juice.
 
Originally posted by: GuitarDaddy
Originally posted by: bap4201
there is a link i posted for 1003-006 beta from germany site. Have you tried/r u using this?

No I've downloaded it, but haven't flashed it yet. I noticed its a zip file, do you just unzip, then flash the resulting file?

guitardaddy, yes it's a zip file. the name was something like 1003sd.006 and it was listed as a ".006" file.

I changed/renamed it to : 1003-006.bin, which made it a bin file. I did this because I've never flashed that type of file , it would probably have worked, but I did not want to take any chances and changed it to the standard bin type file.

Bryan
 
Hey Guys! Here's what I've got so far Prime95 stable for 4 hours:

no particular order:

mem 2.5-3-3-10-13-17 1T ; 2.8v ; ddr400 1:1

cpu 234 x 12 = 2,808 ; 1.6375v


What should be a good cpu temp when running prime 95 for any significant amt of time? Anyone have an average?

I'm running @ 54c watercooled that seems high to me (according to asus probe).

Thanks!
 
Anything under 60C is fine.

Wow 2.8 is very nice - FX-55s don't have a ton of headroom to begin with, so you've squeezed some nice extra juice out of it.

 
Originally posted by: GuitarDaddy
Way to go Bap! That must be one fast rig. Whats your Super PI, 3dmark05 scores now?

I got 30,874 in 3dmark01; 10,010 in 3dmark05; 1M digits of pi = 31 seconds.

video oc = 400/1.06

PS: This OC of mobo was easy to acheive. In fact it was my first try. I simply went into bios first and set ddr = 400 to run 1:1 and oc to manual. Then started up ai-booster and just increased fsb a few points. Exited, restarted, went into bios and set above settings...bam...done. The only other thing that I did was went back in after some testing and set row cycle time = 13 and refresh cycle =15. I going to try to lower them to like 9 and 13.

cpu temp = 39c idle ; 54c when running prime 95 or extreme load, ambient 68F.

and I'm using beta 1003-006

PPS: i guess i can't complain about the 3d05 score, and i don't even have ultras....yet!

Bryan
 
Your temps are a little high bap4201, expecially for water-cooled.

On a lapped XP-120, I overclock to 2.55 GHz and get idle 32C and Prime95 43C in a 75F room.

Seems like the 1003 betas are a little better in terms of stable BIOS settings.
 
Again, 60C is considered the "desirable limit" by OCers, and the chip is probably safe to 70C or so...54C is nothing to worry about.
 
Back
Top