The Phenom II List of Overclocks

[Oldguy932]

My evap head sits at -36C at load during blend. CPU load temp is below 0C, I just don't get any readings below 0 for it. Socket has gone to 9C max at load with 1.5875v vcore and 1.5v set for the nb. I think 1.45v got me to 3.8ghz, 1ghz less than they got. Do you know how low they set everything else when they hit 4.8ghz? I have my ht link set to 7x because its not really necessary. Maybe its also that almost everyone here uses 790FX boards and 790GX might actually be better for it to clock well. I expected a batch of 154 to be decent at least, but thats life.

Edit: Just booted at 3.25ghz NB!!!

[Aquineas]

This CPU is puzzling me, though admittedly I'm no OC'ing stud. I'm seeing 3.5ghz at stock voltage on air. I didn't write down the full batch number but I did note that the 5 digit number from the last line was 80034. I started playing around with voltages and speeds and whatnot but I just couldn't get Vista 64 to boot and I don't have the patience to spend too much time with this. I'm rock solid at 3.5ghz at stock voltage which isn't too bad, so I'm happy..

BTW ACC is set to auto and my machine posts fine, though I'm not sure if it's doing anything one way or another. I know some people weren't getting a POST with it turned on.

[roofsniper]

nice im aiming for 3.5ghz on stock voltage. im most likely going to try and get the highest clocks i can on my system on stock voltage.

[Oldguy932]

That is an amazing batch number mate. Hopefully its a batch that it applies to clocking well.

[jcool]

Oldguy is that a modded Vapochill you use? Just wondering because a stock unit wouldn't hold a Quad at 1,58? Or does the Phenom II take so little power?

[roofsniper]

That is an amazing batch number mate. Hopefully its a batch that it applies to clocking well.

does the batch number really matter? i don't really know much about them but i assumed the only use for consumers and batch numbers were that if you had similar batch numbers then your chips might clock similar.

[Aquineas]

That is an amazing batch number mate. Hopefully its a batch that it applies to clocking well.

Thanks mate. I regret that my camera wasn't charged to get a photo but I was too impatient.


Roofsniper, no idea if it makes a difference but this is an interesting read. I don't know if it's true, but hey, I'm happy. Now to finally install the Fallout patch :-)

[chew*]

My evap head sits at -36C at load during blend. CPU load temp is below 0C, I just don't get any readings below 0 for it. Socket has gone to 9C max at load with 1.5875v vcore and 1.5v set for the nb. I think 1.45v got me to 3.8ghz, 1ghz less than they got. Do you know how low they set everything else when they hit 4.8ghz? I have my ht link set to 7x because its not really necessary. Maybe its also that almost everyone here uses 790FX boards and 790GX might actually be better for it to clock well. I expected a batch of 154 to be decent at least, but thats life.

Edit: Just booted at 3.25ghz NB!!!

Old it was mutiplier only and in AOD, there is another possibility, you happen to have a copy of ETQW? it was able to run that no prime though as it wasnt installed.

As far as batch i'm guessing aquneas chip is a week 50 It appears the average i'm seein in this thread is 3.7 so its possible it will do a little better than average ( as always it also depends on whos ocing the chip ), I was kinda hoping the first batch would come out with a bang.That would mean average speeds in later batches being even better.

[chew*]

does the batch number really matter? i don't really know much about them but i assumed the only use for consumers and batch numbers were that if you had similar batch numbers then your chips might clock similar.

Roof do you really want me to support my claims?

I can tell you this if i posted it I would be giving the intel guys some serious ammunition and I don't think any of us really want that.

[Oldguy932]

I wouldn't even be just windows stable at 1.45v for 4.8ghz. 3.25ghz on the nb won't get blend stable for all 4 cores, a random core always fails with it sadly. At least it runs benchmarks pretty well.

[roofsniper]

yea i read chew*'s thread about the batch numbers. knopflerbruce was correct when he said the L represents the 12th month, the 8 represents 2008 and the last 4 are a wafer id. im just confused as to why they would have 2 different numbers/letters on there telling what the month and years are.

[chew*]

yea i read chew*'s thread about the batch numbers. knopflerbruce was correct when he said the L represents the 12th month, the 8 represents 2008 and the last 4 are a wafer id. im just confused as to why they would have 2 different numbers/letters on there telling what the month and years are.

The bottom line is like a serial # for rma purposes I believe. If you look at lets say a TV for instance it will say manufactured june 08 and then it will have a serial # also including the manufacturing date in the form of a code.

Thats why posting a chip with just the bottom line blotched out is cool.

[charged3800z24]

Phenom II x4 940
CACVC AC 0850BPMW
4.1GHZ
1.55v
2200mhz 1.3v
2x 2GB OCZ 1066mhz 2.1v
UnGanged
ASUS M3A79-T Deluxe Bios 0602
Swiftech H20 Apogee GT
Windows XP Pro 32bit

[charged3800z24]

My evap head sits at -36C at load during blend. CPU load temp is below 0C, I just don't get any readings below 0 for it. Socket has gone to 9C max at load with 1.5875v vcore and 1.5v set for the nb. I think 1.45v got me to 3.8ghz, 1ghz less than they got. Do you know how low they set everything else when they hit 4.8ghz? I have my ht link set to 7x because its not really necessary. Maybe its also that almost everyone here uses 790FX boards and 790GX might actually be better for it to clock well. I expected a batch of 154 to be decent at least, but thats life.

Edit: Just booted at 3.25ghz NB!!!

nice...run RAR on that thing..

[chew*]

Hey Old guy can you do me a favor, I was sleeping and woke out of a dead sleep to type this.......

Run a bench lets say cinebench at 3.0 gig. Drop the HT speed only, down to lets say 1.0 spec......Run it again. See how much it loses.

Then clock the cpu via multiplier only 100mhz and run it one more time.

I think you will find CLOCK Speed > HT Speed.

Once you do all that go after MAX clockspeed with HT under clocked and let me know what the max of your CPU is

Might even want to try underclocking your NB as the next step if the results pan out the way I think they will.....

[NCspecV81]

my stepping is CACVC AC 0850BPDW

[Daveburt714]

Awesome clocks Charged! I'm just curious though, do you need 1.55v's to get those or did you just set it high to eliminate any problems?

I'm only asking because it seems like mine will run much higher clocks at lower v's until my w/c system starts building up heat. If I can keep load under 40c I get better speeds at lower volts.

It's still pretty early with PhII, but it seems to me like it would rather stay cool than have high v's (of course for hardcore OC you'd probably need both )...

I'm going to pick up a Scythe "Ultra Kaze 3000" this weekend and see if it will help keep the heat loading down...

[Lastviking]

does the batch number really matter? i don't really know much about them but i assumed the only use for consumers and batch numbers were that if you had similar batch numbers then your chips might clock similar.

They does not matter much....its more about luck, i got 0058 and its seems not to be one of the best ones. But 24/7 3.8ghz+ looks posible on air.

[jonspd]

Question?

should 2.8-3ghz be possible on the nb speed on a sb600 chipset seeing as the NB is on-die correct?

I haven't pushed over 2.2 atm and was just wondering before I did more test this weekend...

[bldegle2]

i got two phenoms, one is 0007 (supposedly better sillycon), the other 0401, both clock virtually the same, but seem to hit a wall @3.8gig, sometimes I can get more....now, i am using the 790fx-m2r sb600, which may be limiting a bit the clockage and the recent 12/26 beta bios for the board still has some 'issues', not to mention the PWM seems to now be running hot hot hot, the fingertip test was a bit painfull (had to add a fan, after shutting her down when she hit 85*c), which it never did with duallie's....not even close....

seems a stress run triggers the PWM to spike, then it takes awhile for it to 'settle down'....i strongly suspect this is having a huge effect on the clockage and future usefullness of this board.....

i am a bit disappointed in the overall clockage, seems from what i have read that 3.6 to 3.8gig max is the most common clockage, stable being defined by individual users, certainly not near the overall 'hype' projected by AMD at the three 'parties' for the Phenom demonstration......maybe a SB750 chipset board might perform a bit better, but i am not convinced of this yet, at least with the phenom II.....

laterzzzzz..................

[Oldguy932]

HT link is already running on the 6x multi to try and make 4.11ghz stable. It primed for 2:10 but then one core got an error so I upped the nb voltage a little and I'm running 32m. I'll give super low ht link and nb in a bit, can't do much since I have classes today.

[Lastviking]

i got two phenoms, one is 0007 (supposedly better sillycon), the other 0401, both clock virtually the same, but seem to hit a wall @3.8gig, sometimes I can get more....now, i am using the 790fx-m2r sb600, which may be limiting a bit the clockage and the recent 12/26 beta bios for the board still has some 'issues', not to mention the PWM seems to now be running hot hot hot, the fingertip test was a bit painfull (had to add a fan, after shutting her down when she hit 85*c), which it never did with duallie's....not even close....

seems a stress run triggers the PWM to spike, then it takes awhile for it to 'settle down'....i strongly suspect this is having a huge effect on the clockage and future usefullness of this board.....

i am a bit disappointed in the overall clockage, seems from what i have read that 3.6 to 3.8gig max is the most common clockage, stable being defined by individual users, certainly not near the overall 'hype' projected by AMD at the three 'parties' for the Phenom demonstration......maybe a SB750 chipset board might perform a bit better, but i am not convinced of this yet, at least with the phenom II.....

laterzzzzz..................

3.6-3.9 sound about right with air 24/7... abit disappointed here to, the temps seems much higher on retail chips with air... almost like Ci7. Mine is about 3740mhz 1.47v 24/7 temp 38c idle/load 60c+ , Nb 2700mhz+

[charged3800z24]

Awesome clocks Charged! I'm just curious though, do you need 1.55v's to get those or did you just set it high to eliminate any problems?

I'm only asking because it seems like mine will run much higher clocks at lower v's until my w/c system starts building up heat. If I can keep load under 40c I get better speeds at lower volts.

It's still pretty early with PhII, but it seems to me like it would rather stay cool than have high v's (of course for hardcore OC you'd probably need both )...

I'm going to pick up a Scythe "Ultra Kaze 3000" this weekend and see if it will help keep the heat loading down...

I had to run the voltage to get it to run that speed. I can run 3.6ghz/2600NB stock v's, plus 3.8ghz 1.4v/2800NB 1.4v with no issues at all. Any thing higher needs the voltge. Just food for thought. I will aim for higher clocks today. I was ab;e to get my 3000mhz NB back up on the asus..which was where I started on my first attempt a whioe back but killed the board...I think I found it was ACC that did it.

[charged3800z24]

3.6-3.9 sound about right with air 24/7... abit disappointed here to, the temps seems much higher on retail chips with air... almost like Ci7. Mine is about 3740mhz 1.47v 24/7 temp 38c idle/load 60c+ , Nb 2700mhz+

I noticed Higher NB and of course the extra v's to get it there brings up the temps.

[Lastviking]

I noticed Higher NB and of course the extra v's to get it there brings up the temps.

Yeah higher voltage give alot extra temps.... but i dont see any higher temps becuse of higher NB here...yes the pcw and system get hoter but not much, main problem is to get the cpu temp down....i think i trow the stuff out in the snow soon