Conroe CPU and CPU core temp comparison thread

[phile]

I'd like for people to post their temps as read from the mobo diode (Probe II, speedfan, etc.) and from the two core diodes. (Everest Ultimate is the only program I know of that will read all 3 - its core readings match perfectly those of CoreTemp.)

I'm getting an approximate 20C delta T between the mobo CPU diode and the core diodes, and find that very odd. It makes complete sense to me that the core temps will be hotter, as those diodes are in a hotter location than the mobo's cpu diode... but 20C? That seems excessive. Please include which conroe you have, your mobo with bios version, and list the clock speed and vcore.

E6600 @ 2880
P5W DH Deluxe w/ 1101
vcore auto

Idle CPU temp (mobo): 31C
Idle CPU core1: 50C
Idle CPU core2: 51C

Note: My core2 is always one ore two degrees C higher than core1.

Thanks.

-phil

 

[renozi]

E6700@ 2671
P5W DH Deluxe 1101
vcore auto

Idle CPU temp (mobo): 31C
Idle CPU core1: 41C
Idle CPU core2: 42C

I guess you are right Phil. Core 2 is always higher than Core 1.

 

[aggressor]

I don't have idle scores since I'm in the middle of a Prime95 stress test

E6400 @ 3.2ghz
Asus P5B-Deluxe BIOS 507
Vcore: 1.344v actual
Core #1: 75C
Core #2: 74C
Motherboard: 47C

Coretemp reports 79C for both cores, so it's slightly different from Everest.

 

[renozi]

I don't get how people are overclocking so high with lower voltage. I have to go to 1.425v to get my E6700 to stable at 3.33Ghz. with Mem at 833 4-4-4-12 2.0v (Corsair PC2 6400C4). Something I'm doing wrong?

 

[neuralnut]

My 6400 runs 100% stable at 3.2 GHz on 1.35 Vcore, hasn't needed any more. I have found the limiting factor to be memory and have the VMCH at 1.6 and the Vdimm at 2.1 however.

I think people are assuming that when their systems start to get unstable that they need to give the CPU more volts when the limiting factor is the memory side of things. Try getting your system stable then progressively lowering Vcore, you may be surprised how much you can lower it before the system gets unstable

 

[Janne2]

E6300 3220mhz 1.3v
Easytune 44c
CoreTemp 54-55
Everest 54-55
Systool 70
DS3 bios F4
Zalman 7700

 

[FallenHero]

E6600 @ 3.2 1.45vcore
P5W DH Deluxe w/1101

Core temp is 37C idle, 45 under normal load conditions

 

[phile]

FallenHero

Is that 37C the temperature read from the mobo CPU diode, or from either of the two core diodes? That was the purpose of this thread; to observe the delta between the two zones.

Thanks.

-phil

 

[tylerw13]

e6600 @ 3.3 1.45 vcore
p5wdh deluxe 1101

cpu 36c
core 1 50
core 2 51
mb 40

 

[GuitarDaddy]

Originally posted by: aggressor
I don't have idle scores since I'm in the middle of a Prime95 stress test

E6400 @ 3.2ghz
Asus P5B-Deluxe BIOS 507
Vcore: 1.344v actual
Core #1: 75C
Core #2: 74C
Motherboard: 47C

Coretemp reports 79C for both cores, so it's slightly different from Everest.

What HSF? Your temps look pretty high

 

[aggressor]

Yep, they do, but I'm not sure if they are accurate or not. A lot of people with P5B Deluxes are getting high temps like that.
I have an XP90 for cooling.

 

[imported_drummerboy]

E6700 @ 2671MHz
P5W DH Deluxe 0801 BIOS
vcore auto

Idle mobo cpu temp: 28C
Idle CPU core temp 1: 44C
Idle CPU core temp 2: 43C

 

[serializable]

MSI 975x Platinum with 715 BIOS
E6600 @ 2400
vcore auto

Idle Core #1 34
Idle Core #2 29
Idle Mobo 31

I don't know why there is such a big difference between the 2 cores. Everyone elses is only like 1 degree difference.

 

[kpamir]

p5w dh
e6600
default everything cpu wise atleast

idle 28-29
100% load both cores 38-39

 

[phile]

Seems to me like the CPU core temps vary significantly from machine to machine, while the CPU mobo temps fall with a much smaller range. My motivation to poll everyone on these temps was motivated by a slight distrust of these core readings, and the results so far have served to justify that distrust.

-phil

 

[kenograndmaster]

E6600 3469 , 385x9 1.45v, air (scythe ninja plus)
P5WDH bios 1101 (801 same thing)

Speedfan, Probe,Everest Home: idle:43
load:55
Coretemp Beta, TAT : idle: 55
load: 74

room temp: about 77c <---( SORRY EDIT. I mean 77F so about 25-26C)


NOTE: weird thing is that these load temps are from running Rosetta@Home but when i run
Prime95 (2 instances, Small FFT), my core temp load is 83 and speedfan/probe/everest load
temp is 65-67.

Ive read Rosetta is a better stability test so im going to stick with rosetta but it seems when im at 100 percent cpu load, my load temps are lower than prime95. Both load temps are scary but prime95 is just too high. Ive tried to reapply Arctic Silver 5 and remount but same thing.

What do you guys think? Should i lower my voltage down so that my core temp load is under 61 or leave it alone.

I think Core Temp is a little higher than what it should be, i know its diode temp but when i touch my Heatsink its not really hot, its slightly warm. Does that mean my cpu at 74 load (core temp, 1.45v) is accurate? I know it should be more than Probe/speedfan but its a delta of 19 degrees give or take?

Let me know what you guys think about that. Im all ears

-keno

 

[imported_Zeke]

Originally posted by: kenograndmaster
E6600 3469 , 385x9 1.45v, air (scythe ninja plus)
P5WDH bios 1101 (801 same thing)

Speedfan, Probe,Everest Home: idle:43
load:55
Coretemp Beta, TAT : idle: 55
load: 74

room temp: about 77C


NOTE: weird thing is that these load temps are from running Rosetta@Home but when i run
Prime95 (2 instances, Small FFT), my core temp load is 83 and speedfan/probe/everest load
temp is 65-67.

Ive read Rosetta is a better stability test so im going to stick with rosetta but it seems when im at 100 percent cpu load, my load temps are lower than prime95. Both load temps are scary but prime95 is just too high. Ive tried to reapply Arctic Silver 5 and remount but same thing.

What do you guys think? Should i lower my voltage down so that my core temp load is under 61 or leave it alone.

I think Core Temp is a little higher than what it should be, i know its diode temp but when i touch my Heatsink its not really hot, its slightly warm. Does that mean my cpu at 74 load (core temp, 1.45v) is accurate? I know it should be more than Probe/speedfan but its a delta of 19 degrees give or take?

Let me know what you guys think about that. Im all ears

-keno

Its 77C in your room? Holy crap. How are you still alive?

 

[phile]

kenograndmaster,

I think you need bring down that voltage, and thus your o/c. I have my E6600 @ 333FSB w/ 1.35v, and that's as high as I'm going to take it, on air. My temps start to shoot-up when I take the vcore higher than 1.35. I magine at 1.45v my temps would be where yours are.

Current idle temps:

CPU (modo diode): 31C
CPU core1: 51C
CPU core2: 52C
Mobo: 37C
Cooler: Zalman CNPS9500 LED @ 2400rpm

Ambient room temp: around 20C

My cores hit the mid-60s at load, which is about as high as I'm willing to let them go.

-phil

 

[Ryan T]

E6400 @ 2.67 GHz , stock voltage

idle 37 degrees
Both cores 100% stress test: 54 degrees

If I go any higher than this, the system becomes unstable. If I wanted to take this bad boy up to say 3 GHz, what should my voltage be?

 

[phile]

I just found a great way to lower temps for those of you not interested in overclocking: under-volting. On my ASUS P5W DH vcore at auto delivers roughly 1.32v to the cpu. I am able, however, to lower the vcore all the way down to 1.25, so I did so. I ran dual prime for an hour, pcmark05, 3dmark06, 3dmark05, and some memory benchmarks, all without issue or degradation in performance. Awesome. So, how does this affect temps? Significantly!

Check out these idle temps for my E6600/P5W DH Deluxe:
(FSB 266, vcore 1.25)

CPU (mobo sensor): 26C
CPU core1: 42C
CPU core2: 42C
MOBO: 36
Cooler: Zalman CNPS9500

That's a 10C drop on the core temps.

-phil

 

[phile]

Originally posted by: Ryan T
E6400 @ 2.67 GHz , stock voltage

idle 37 degrees
Both cores 100% stress test: 54 degrees

If I go any higher than this, the system becomes unstable. If I wanted to take this bad boy up to say 3 GHz, what should my voltage be?

Start at 1.35v and run dual prime. It should fail very quickly if it needs more juice. Go back and try the next step up, and dual prime again. Rinse repeat.

-phil

 

[lutorm]

E6400 rev B2, DFI Infinity 975X/G. Stock speeds and voltages. Scythe Ninja heat sink. Ambient temp 23C.

Idle:
CPU temp (speedfan, etc all show consistent temps): 36C
Core temp (core temp beta 0.9.0.91): 36C / 40C
Heat sink base (DS18S20 temp probe): 33C

2xPrime 95 In-place FFT:
CPU temp (speedfan): 55C
Core temp: 55C / 56C
HS base: 34.5C

Interestingly my core and mobo temps remain consistent from idle to full load, unlike what many others have reported.

On the other hand, the deltaT from CPU to heat sink base goes from 3C to 21C. This seems very high to me, at TDP (which I'm almost certainly not at) this means a thermal resistance of 0.3C/W just between the die and the heat sink base. This is a higher thermal resistance than that from die to AIR in the Ninja review at SPCR (http://www.silentpcreview.com/article251-page4.html). There must be something wrong with the seating or the AS5... How long does it take for the AS5 to "cure" and how much difference does that make?

 

[Polish3d]

Stepping 6 Rev B2 E6600 Week 26


3.57mhz, 1.44v, high end air (Big typhoon w/ 3 120mm fans via antec p180)


CoreTemp:

Idle: ~ 40c

Dual Prime load: Up to 58c

 

[Kyanzes]

E6300 @ 2800
Intel 975XBX 1334 (w/ the small hw mod)
vcore auto
Stepping: 6/F/B2

Idle CPU temp (mobo): 39C
Idle CPU core1: 37C
Idle CPU core2: 41C

HSF: 9500LED on full RPM (2500revs)

Mobo can't handle much more FSB it becomes unstable at around 410MHz

 

[kenograndmaster]

I use P5WDH e6600 360x9 at 3.2 1.35v

1.35v on scythe ninja air with room temp 25C

core temp is idle: 53 speedfan/probe/everest: 38
core temp load: 74 " " " : 55-57 (that means at 1.35 which is almost close to stock is 4 degrees away from max thermal temp for these conroes at 61C which
seems really fishy)


Ive seen people running 1.4 volts air with about same room temp with scythe ninja as well
as other high quality heatsinks yet their core temp is lower or higher so im thinking this
core temp is wrong because if it was right then ppl running 1.4+ volts should have temps above 61 load.

When i get my chipset coolers i will remount the HS just to make sure the contact is there but im sure it is because other ppl have tried remounting many times yet still same temps.