Intel cpus at UVOC, post & share your result here !

[bing]

Torin,

Wow ! Just knew it if that mobo can go as low as 0.4 Volt ! Although I'm sure at that level it will be usable with current CPU generation.

For UVOC test, the VCore should be manually set and don't use Auto.

Just use the VCore setting (not the reading) from the bios, and again not from the CPUZ.
Fyi, nowdays the vcore set by the bios is quite accurate, especially for recent highend OC-ing mobo starting from P965 generation.

Yes, you're right, CPUZ vcore reading is a hit and miss situations at different mobo models and different Super IO chips, and until there is a unified method and initiative from Super IO chip maker and/or Intel, Franckz (CPUZ author) will never able to put the VCore value in the CPUZ validation.

Just do the UVOC method as I mentioned at the 1st post, and I'm really curious for your temp delta of yours since you're on highend water.

Thanks for the offer, dunno if its useful, you could try UVUC (Under Voltage Under Clocking) .. LOL ..sounds funny.
It's the way to find the "lowest" cpu speed by lowering the multi or/and fsb with the "lowest" VCore as possible.

Can't imagine the temperature delta at UVUC, must be really-really low !

Regarding your upcoming UVOC result, just put it on a new post instead of editing the old post.

 

[OneArmedScissor]

Just got my Xeon X3210 G0 in and right now under full load (stock settings, etc.) my Vcore is 1.264.

 

[bing]

OAS, thx, but I think that is OVOC, have you read the requirement and detail for UVOC at the 1st post ?

Expecting & waiting for your uvoc result here !

 

[OneArmedScissor]

OAS, thx, but I think that is OVOC, have you read the requirement and detail for UVOC at the 1st post ?

Expecting & waiting for your uvoc result here !

Oh I know!
I was giving you guys a heads up and a "control," so to say, before the overclocking fun begins!

 

[bing]

Great ! Second Xeon X3210 G0 entry in the table is coming then !

 

[torin3]

For UVOC test, the VCore should be manually set and don't use Auto.

Understood on the Vcore, but I was asking about the Auto on the Startup option. I've got quite a range, but I figured Auto would probably be more representative of what the board is doing as it is one less option I'm fiddling with. I'm going to try the 1.02500 tonight with the startup on Auto and see if that crashes as well.

 

[torin3]

Well, so far I'm up to 20 minutes of prime95ing with no problems at 1.02500v.

Also, the idle temps were averaging 39°C, and under load it is averaging 45°C.

If this is still priming when I get up in about 4-5 hours, I'll start overclocking it at this voltage.

Also, the lowest Vcore is 0.44375v, and the highest is 1.60000v.

Also there is an oddity in the Vcore settings. The voltages seem to be in an even progression until you hit 1.0000v, then the settings, in order are:
1.00000v
1.06255v
1.06255v
1.01250v
1.01875v

Very strange

 

[torin3]

Well, so far I'm up to 20 minutes of prime95ing with no problems at 1.02500v.

Well it lasted for about an hour. Auto on Startup % seems to have no effect. Oh well, next step up is 1.03125v.

 

[bing]

Wow ! 6 Celcius of delta temp, that is a really good cooling you got there !

Regarding VCore setting from your mobo, yeah, its very strange sequences.

Fyi, these are the official Intel's Q6600 VIDs list and they're in 7 bits notation.

0.8500  0.9125 1.0125 1.1125 1.2125 1.3125 1.4125 1.5125
0.8625  0.9250 1.0250 1.1250 1.2250 1.3250 1.4250 1.5250
0.8750  0.9375 1.0375 1.1375 1.2375 1.3375 1.4375 1.5375
0.8875  0.9500 1.0500 1.1500 1.2500 1.3500 1.4500 1.5500
0.9000  0.9625 1.0625 1.1625 1.2625 1.3625 1.4625 1.5625
        0.9750 1.0750 1.1750 1.2750 1.3750 1.4750 1.5750
        0.9875 1.0875 1.1875 1.2875 1.3875 1.4875 1.5875
        1.0000 1.1000 1.2000 1.3000 1.4000 1.5000 1.6000

Looks like that mobo is capable of finer resolution on VCore selection which is a good thing.

 

[torin3]

Wow ! 6 Celcius of delta temp, that is a really good cooling you got there !

Thanks, I was a little worried about maybe having applied too much of the diamond TIM. It is warmer today, about 31°C ambient, so the temps are higher, but the delta seems to be the same. 41°C idle and 47°C under load.

Currently at about 20 minutes into this run at 1.03125v

Edit: Well, it lasted about 2 hours. Didn't blue screen, and as far as I could tell, prime never crashed, but the taskbar got messed up and the start button would have less and less I could click on. Logged the account out and then back in without rebooting, and everything looked good, but I would say that that incident would mean unstable. Up another voltage notch then.

 

[torin3]

Ok, lowest stable Vcore is 1.03750v

Primed for 3 hours, and hammered it with running multiple applications and just trying to load it as much as possible.

Now trying to see what type of overclock it will do at this voltage.

2.475Ghz/1100FSB lasted 5 minutes and blue screened.
2.452Ghz/1090FSB lasted 15 minutes and locked up.
2.441Ghz/1085FSB still running a little over 15 minutes at this point. If it lasts about an hour I'll start hammering it on top of the prime.

Edit: 2.441Ghz locked up at the 1/2hr point and reset the BIOS settings too.
Edit:
Edit: Now testing:
Edit: 2.430Ghz/1080FSB.


Idle 40°C / Load 46°C

Ambient is about 28-30°C

 

[bing]

Another UVOC result in making, appreciate the effort, time spent and the patience.

 

[torin3]

Another UVOC result in making, appreciate the effort, time spent and the patience.

Hey, glad to help out. I admit I was curious too. However, my results aren't quite as good as the other DFI 680i board listed in the grid.

Max stable overclock at 1.03750v:

2.419Ghz (only 19 Mhz)/1075 FSB

Stability checked with Prime95, ran with no errors for 8 hours.

With ambient temps around 28-30°C Idle temp was 40°C, load was 46°C

Any additional details you need?

 

[bing]

Torin, thank you sir ! Especially on the 8 hours test !

All your details is completed, check the table to verify your entry !

One interesting fact, your delta temp is the best so far, 6 degrees Celcius from idle to load !

This proof that your mounting at your wc block to the IHS is at their best condition, and also your loop is performing really good, congrats !


To all :

Anyone want to jump in with some conclusions for data samples gathered so far ?

Me 1st :

• Delta temp on UVOC should no more than 10 Celcius, especially you're not using stock Intel HSF. (This needs to be verified with owner who use stock HSF on UVOC)
• From the results, E6x50 needs more vcore (avg = 1.2) than Q6600 G0 (avg = 1.0) at stock speed, hence it is safe to conclude that E6x50 family has "tighter" binning between E6X50 themself while quad G0 are more relaxed, this fact is also supported by the (VID - Lowest VCore) average value.

Please, feel free to drop some comments or better other conclusions that you have in mind after reading the table.TIA.

 

[smilingcrow]

From the results, E6x50 needs more vcore (avg = 1.2) than Q6600 G0 (avg = 1.0) at stock speed.

I ignored the G0 tests that were run for less than an hour or were not at stock speed and the average came to 1.0375V.
My E2160 (L2) at 1.8GHz requires 1.025V, so 1.0375V at 2.4GHz is sweet.
I’m going to test my E4400 (L2) at 2.4GHz and will post the results in case anyone’s interested!

 

[torin3]

Torin, thank you sir ! Especially on the 8 hours test !

All your details is completed, check the table to verify your entry !

One interesting fact, your delta temp is the best so far, 6 degrees Celcius from idle to load !

This proofs your the mounting on your wc block to the IHS is at their best condition and also your loop is performing really good, congrats !

Thanks! It all looks good.

I do notice that my delta is pretty close to the other Apogee user's delta and I'm wondering if it might be the diamond TIM making the difference.

I'm very glad to see it that it looks like I installed it right.

That was actually kind of fun. Let me know if you want any other stability tests done.

 

[bing]

I ignored the G0 tests that were run for less than an hour or were not at stock speed and the average came to 1.0375V.
My E2160 (L2) at 1.8GHz requires 1.025V, so 1.0375V at 2.4GHz is sweet.
I’m going to test my E4400 (L2) at 2.4GHz and will post the results in case anyone’s interested!

Hi Smilingcrow,

You make me think and consider to change this thread's title to be more suitable for wider audience, thanks !

And YESSSS ! Any UVOC results are wellcome, and please do it both on your E2160 and E4400 if you don't mind.

Please read the details required in the 1st post, thanks for your offer !

Thanks! It all looks good.

I do notice that my delta is pretty close to the other Apogee user's delta and I'm wondering if it might be the diamond TIM making the difference.

I'm very glad to see it that it looks like I installed it right.

That was actually kind of fun. Let me know if you want any other stability tests done.

Yes, you could be right, but again, it needs others more samples to confirm these since one is not enough imo, and also the difference is only 1 degree, could be in the range on error margin.

Anyway, it must feels good to know that your cooling is working at it's best performance !

So far I think it is enough for the stability test, that 8 hours means your result is really reliable.
Thank you very much for your offer, really appreciate it.

 

[Maverick0984]

I plan on finally getting around to doing this with my Q6600 tonight, or worst case this weekend.

Only problem is I have a P5K Premium which looks to have the 1.1v minimum. Oh well.

Question though, for CPU temp do you want an average of the Cores? Or the temp labeled CPU in something like Everest or Speedfan?

 

[Maverick0984]

My Q6600 G0:

CPU: Q6600
VID: 1.25v
Lowest Vcore: 1.1v (P5K Premium restriction)
FSB at 1.1v: 305Mhz
Multiplier: 9x
Clock Speed: 305*9 = 2.745Ghz
Stabilitiy Test: Prime95 Quad core version (Small FFT NOT Blend) for 1 hour
Idle/Load/Ambiet: 30C/39C/27C
Mobo/Bios: P5K Premium w/ 0204 Bios
Cooling: Ultra 120 Ex
FPO/Batch: L723A841
Pack Date: 07/27/2007

 

[bing]

Maverick, thank you !

Yeah, too bad Asus P5K mobo series with minimum VCore limitation at 1.1000 Volt, I put them at different background color at the table and with a remark about it !

Your temp delta at 9 degrees is also looks ok.

Table updated with your entry, again, thanks !




.