How well does your Opteron overclock?

I do believe those are one of the good ones! Good job bro...only 1 way to find out

CPU Type: Single 146
Stepping: CABYE 0540 FPBW
Best Stable OC: 3000ghz
VCore: 1.47v
Temps: idle 34C/ load 49-50C
maxcaseT: 63C

The processor will go faster on that voltage, but the temperatures are to high at the current time to go faster. I have not burned in at all.

That's an awesome voltage to hit 3.0 at! You seem to be running on the hot side. What are you using for cooling? After burning in for a few days, my load temps dropped to 42C @ 1.550v. I would imagine that your chip could be generating even less heat than mine.

It looks like the 144 and 146's are on avarage doing way better then 148's. I feel so cheeted. I was getting a 146 but I got there a day to late and there where only 148's left.

ocZer said:

It looks like the 144 and 146's are on avarage doing way better then 148's. I feel so cheeted. I was getting a 146 but I got there a day to late and there where only 148's left.

Click to expand...

You might want to increase vcore, or try different RAM dividers if you can. Or, try a different multi. I know there are some multis that my 148 doesn't like as much.

I have tried all VCore from 1.4 to 1.525 of wich 1.5 was the one that gave me the worst results. Temps are just to high until I get better cooling. When it comes to multipliers, I have yet to try anything other then 11x, but I will. Pacience young padawan, pacience.

Added links for the above post...grrrrr.
Screen 1
Screen 2
Screen 3

ocZer said:

I have tried all VCore from 1.4 to 1.525 of wich 1.5 was the one that gave me the worst results. Temps are just to high until I get better cooling. When it comes to multipliers, I have yet to try anything other then 11x, but I will. Pacience young padawan, pacience.

Click to expand...

Oh ok, well, once you get the heat thing under control, I think you'll start to see much better results. These are surprisingly sensitive to heat. Anything close to or over 50C has the potential to cause errors.

SN: Saw that you posted me under the 175's... I got a 170 man Don't want to throw your results off.

Also its a RPMW, just realized that i typed that wrong in my previous post doh.

CPU Type: 170
Stepping: ccbwe 0530tpmw
Best Stable OC: 2.6ghz
VCore: 1.4
Temps: 29/55

I'm sure it can go higher but i don't want to try because of the high temps.

Yeah, that's pretty hot. The Zalman7000 is awesome and super quiet, but not the best for overclocking...

My burn in has been done sloppily. I pretty much did it the way Nade told me until I started getting a lot of crashes at around 2650 the first day I really did it. It was more of a thing that I was choosing to do and bend the rules a little. I was trying to get the absolute minimum stable vcore to burn it at and I began to have problems around 2650 because 2590 is where my 1.30 needs to be upped to 1.325+. So I must have crashed and rebooted a dozen times between 2650 and my target of 2750 during the day I was initially burning-in.
So I can not say my work was well done, I never kept notes and everytime my computer crashed I rebooted as quickly as possible at the next higher voltage, burn in for a few hours and try to lower the volts back to the crash level and burn more (hopfully with no burn-in interrruption). I was eventually able to reduce and seemingly eliminate those crashes once the burn-in ran for a few hours, then it would be lowered to the crash vcore. But I feel it's more of a warming-up the computer effect as I've noticed that it tends to be more stable whenever I have it burning in for a few hours.
As a result, I've have been able to reduce vcore down as much as 0.5 to 0.6 volts settings that it would normally crash if I were to just jump in and start burning at that very unstable low voltage.
However, I can tell you this, depite my sloppy record keeping that I pretty much keep in my head, I am able to run 2750 at 1.425 volts off the bat now instead of 1.45. One example of how low I can go is if I start burn in at 2750 and 1.425, I can eventually get it down to 1.393 after it's been burning uninterrupted for a few hours. Of course after it's cooled down and restarted it just wont be burn stable at 1.40. Failes almost immediately. I do believe his process has done something to enhance the chip: I am getting more stable at the same mhz than before I started. And overall, I've been able to reduce the vcore at least 0.025+ at the same settings. I've never ran my cpu at 2850 with out it having an error in a few seconds. But the last 2 days I've been able to do it up to 40 minutes at 2850 and even a few minutes at 2900. But, I can't say that I really practiced this high frequency stability very much before the burn in. I'm just too lazy to keep notes, take time and write everything. I do feel confident that this method has for sure shaved off at least 0.025 volts and increased the length of time my cpu can be stable at those volts at higher frequencies (2750 and higher). I'm currently working at 2850 and 2900 at 1.5-1.525 and 1.550-1.575 respectively. I think I'm seeing results and am trying to push my core0 more. I am ONLY burning in core0 becuase I just can't imagine what my temps would be if I burnt both cores at the same time. I'm hitting 47-52c with that one core and 2900 hit 55c today. It was too high, had to shut it down. My other core doesn't need the burn in until core0 catches up.

FinallyUpdated mine (well added for the first time), first page

Super Nade: Can you update my current status in the list as 2500 @ 1.46V. Temps are now 27/42 as well! nice to have the heat turned down!

Still working on this though! have some new OCZ platinum 2GB coming in today I hope, when it does I'll start pushing again from this point as well! been simply burning in the CPU's at 2550 to try and get that lazy core down a bit more! II have one core that's prime95 stable at 2500 @ stock voltages! this other one don't quite like it!

Do you think it would work to simply burn in the unstable core and leave the stable core to rest? or should the system be fully loaded to have this burning in work? Just curious as to what people think?

Thanks!

O.

Opteron 148 - Soundx98 - CABYE0543FPBW - 1.5V - 26/48 C - 2900 MHz

Opteron 146 ocs to 2.8ghz in my shuttle sn95g5 w/ 1.475 vcore (psu seems to undervolt a bit and displays as 1.426 in cpu-z. Idle temps 32C Load 42C. Could probably go a bit higher but 280fsb is the highest this shuttle will go.

Updated my results so far..... I'm up to 2.45 ghz on my 144 (270x9) on stock voltage (1.4) and cooling.

146 CAB2E 0540

3.10 @ 1.525V

Watercooled

Removed IHS, although it could do 3.10+ before the removal.

Updated, December 27 2005.

ROH, You could set the bun-in program's affinity to one core and let it burn for about 8 hrs. However, I'd recommend that you stress both cores. The extra burn-in won't hurt the stronger core. That's why I do all burning in when I'm off to school or overnight.

Hope everybody had a great weekend.

Updated my first post
http://www.ocforums.com/showpost.php?p=4139997&postcount=3

For your references, Nade