The Official OC (OverClocking) Thread!

well this puts my temps into perspective.
Soph as you and i have talked alot about my temps and since useing realtemp only my temps arent as bad as i once thought. as i have been thinking,the 45nm chips are a breed of there own and cant be compared to the 65nm.
Mort you have the E3110 basicaly the same as the E8400. granted your at 4.0 and im back to 3.6ghz but when i was at 4.0 for the 3 days your temps were only 3-5c lower than mine. i wish i had them temps, but as Soph pointed out to me it is still well within the normal range

right now my Coretemps being read with realtemp is at 43c, but in everest and occt it is 53c but my cpu temp is 31c. and my low in realtemp was 39c.

 

cincyrob

I am still not quite certain that I would go with realtemp because it tends to disagree with other monitors that have also been updated. I tend to go with Coretemp's measures for now, but if it turns out that readltemp is right then great.

 

Soph,

the point I was trying to get across is that realtemp is more accurate for core temps on the 45nm cpu's. if the cores temps are shown to be 68c with realtemp then they are going to be shown even hotter with other software. using your pc with the 45nm cpu, open up realtemp and core temp and look at the differences they are showing for core temps.

the ambient room temp right now is about 77f and here's what the two proggys are showing for mine at idle:



see what I mean?

 

yes, he used REALTEMP for measuring the CPU temperature and it was 68C...he said RealTemp showed that he was using 1.25v (but i think thats the VID for his CPU and not the actual voltage his CPU is using)...
In other news i got my Q6700 running (finally) and the idle temperature according to Nvidia Monitor and ASUS Probe is 17C!!!!!!! and the Real Temp Temperature for my CPU is 33-34-31-27!!!
under 100% load (2 Instances of Orthos - with one's affinity set to Core 0 and 1 and the other's affinity set to Core 2 and 3)Nvidia Monitor and ASUS Probe show 38C! and Real Temp shows 51 - 51 - 46 - 45...
The VID for my chip is 1.325(WTF!!!!)
-im1992

 

I need to correct myself. I heard that the correct TJ max for 45nm cpu's was 95c as reported by realtemp not 105c as reported by other software.

 

Yes I understood that, but coretemp has also been updated for 45nm processors which kind of skews things a bit. I'm neither agreeing nor disagreeing, because I haven't decided which to believe. I've been using real temp but I'm not fully convinced yet.

 

mort

Your version of coretemp is dated get version .99.

 

Soph,

what are the latest versions of coretemp and realtemp? I may need to update mine. been a month or so since I checked for new versions.

 

ok. disregard my last post

 

mort

Just did a little test and found where the discrepancies lie. Realtemp gives a lower Tjmax reading than coretemp because Coretemp is using a T-j max of 105 while Realtemp is using one of 95. If you set the Tjmax in realtemp to 105 then the measures are the same. The real question is which developer is using the correct Tjmax?

 

ya I know it. I was also trying to point that out earlier in a previous post. the $10,000,000 question is, which one is right?

if my temps were 68c with realcore, I would be very concerned since it shows the core temps to be lowest of any software (about 10c in my case). it would scare me to death to see what coretemp was reporting if realtemp is showing 68c.

 

Although realtemp uses a lower tjmax than coretemp the reality is that the distance between the actual core temp and its tjmax is the same for both so it doesn't matter which is right.

 

ok, I'll rephrase. the $10,000,000 question is what is the correct tjmax?

 

which bios is best for a quad core overclocking on the ASUS Striker Extreme(680i)?
thanks,
-im1992

 

I don't actually know which is the real tjmax. Since most sensors were agreeing with coretemp I took it as probable and it might still be true, but in the end it doesn't matter. The concern is the distance between the actual core temp and its tjmax. A tjmax of 95c with a core temp of 58c is just as close to maxing out as a tjmax of 105 and a core temp of 68c. 95-58=37 and 105-68=37 so no difference it is just as hot either way.

 

lol
my q6700 cant even hit 3.2ghz
2.7 is the max
w/e, i guess my luck just s****
-im1992
EDIT: well, i managed to run it at stock speeds with just 1.25v
is than any good for a quad to run at this little voltage?
-im1992

 

im1992

Trust me your settings are wrong if that is all that you can get. That processor should go to sleep hitting 3.4 GHz. I don't know your bios but I could give some advice if I knew your setup.

 

Well here it is:
ASUS Striker Extreme 680i (BIOS 1303 - heard its the best for quad overclocking)
Core 2 Quad Q6700 @ stock speeds
EVGA GeForce 8800GTS - SLI
OCZ PC6400 (4 x 1gb for a total of 4gigs) RAM
500 Watt PSU
If you need me to post any screen shots of the BIOS or something, i could do that too...
For the testing, i took out all the RAM except for 1 stick and also took out one of the GeForce 8800GTS. And still, i couldn't get it to overlock any.
The voltage setting in my BIOS are (WHERE YOU ARE SUPPOSED TO PICK THE VOLTAGES IN BIOS):
VCore Voltage: 1.25v
Memory Voltage: AUTO
1.2V HT Voltage: AUTO
NB Core Voltage: AUTO
SB Core Voltage: AUTO
CPU VTT Voltage: AUTO
DDRII COntroller Ref Voltage: AUTO
DDRII CHannel A Ref Voltage: AUTO
DDRII CHannel B Ref Voltage: AUTO

The Voltage Monitor shows (IN BIOS):
VCore Voltage: 1.23v
Memory Voltage: 1.88v
1.2V HT Voltage: 1.26v
NB Core Voltage: 1.24v
SB Core Voltage: 1.53v
CPU VTT Voltage: 1.26v
DDR Termination Voltage: 0.92v
3.3v Voltage: 3.28v
5v Voltage: 5.18v
12v Voltage: 11.77v

BTW, the CPU that was in it before (an e6300) was overclocked to 3ghz with everything on AUTO and the CPU voltage at 1.15v running at a 1765 FSB...
thanks for the help,
-im1992

 

I will help more tomorrow but a quick question. Which ocz 6400 do you have? I have to download the manual to review settings

For now:

Set you settings options to manual not auto.

Set frontside bus strap to 266 MHz

Increase you memory voltages to 2V

set PCIe bus to 100

Set (if you have it) PCI bus to 33.3 Mhz

Then try to increase you FSB a little at a time.

 

Man those OCZ Platinum PC2-6400s are HORRIBLE. I have had 3 2gb sets fail in customer builds within one month!

Oh and you might need to increase your Vcore a bit more. I had to put 1.45v in my Q6600 to get 3.6Ghz stable. Of course that was AFTER I did a vdrop mod, before it wasn't even possible to get 3.6Ghz stable(and I was too afraid to push the voltage beyond 1.5v, this was on a GA-P35-DS3R).

P.S. Your voltage on the 12v rail is a bit low isn't it?