It seems generally speaking core temps should be about 15 degrees higher than your cpu temps. When I run TAT my cores never exceed 54C, but my CPU according to nTune and SpeedFan is at 50C. I have an 6600 that I would like to OC but I don't know what temps to trust. Should I stop my OC when TAT reads 60-65 underload or should trust that speedfan and ntune are correct and not oc?
I have excellent airflow in my case my room temp is about 20C and I am using a Scythe Ninja Rev B.
PS My current idle temps are 34 speedfan and ntune and 39 using TAT.
pls help me out guys, under speedfan: temp1 57c, temp2 36c,temp3 -2c, core0 32c, core1 30c. pls take note of temp1 57c indicating a burning figure on my speedfan page. TAT: CPU0 33, CPU1 31. my worries is on temp1, is it normal? even at normal clocking still i get same figure of temp1.pls help me.
GA p35 ds3r
e6750 @ 2.8 ghz
vcore1 1.36v, vcore2 1.81v
512 x 2 ddr2 667 @350 mhz
x1950pro 512 mb/128 bits
2d aurora full tower
stock cooler
help me out guys, my temp1 under speedfan goes as high as 57c idle, the rest are as ff (speedfan): temp2 36c, temp3 -2c, core0 32c, core1 30c.i find it disturbing to look at the burning icon right beside that temp1 57c figure.m sure this is not normal considering its burning.can u pls help me whats might be wrong with this?
GA P35 DS3R
e6750 @ 2.8ghz
2x512 DDR2 800 @350mhz
x1950 pro
stock cooler
3d aurora full tower
First of all let me say I'm spanish and you may found several errors in my english. On the other hand, what a great post about thermal in a C2D. Even though it's hard to be understood due to the technical language, it shows a lot of light to people like me.
Let me explain myself:
I do have an ASUS P5B-Deluxe WiFi mobo with an Intel P965 chipset. I've read there are some problems regarding how this chipset measures temps, so I'll keep this in mind.
My rig as follows:
- Mobo as stated. - Core 2 Duo E6700 (Not OverClocked) - 4 Gbs Kingston Hyper X working at 667Mhz - GeForce 8800 GTS 320Mb
It's summer here in Spain and the room temperature it's close to 30º Celsius.
I'm using Everest Ultimate Edition version 4, so I'm guessing values displayed by this tool are accurate.
When the processor is running @ 1.600 Mhz temperatures displayed are Processor: 52 / Core 1:50 Core 2:50
If the processor goes to full load @ 2.666 Mhz, temperature goes to aprox. 64 /Core 1: 62 Core 2:62-63
Do I have to be worried?
Thanks in advance and thanks for that great post. Apologize any inconvenience and thanks for your time.
My temperatures are reversed. Just running firefox, aim, msn and various other minor software I get these readings from Speed Fan
System:26
CPU: 36
Core0: 23
Core1: 23
TAT measures both cores running between 24-27
If Core temperatures are supposed to be higher than case temperatures this seems to be erroneous.
of all the programs that report temps and voltages, what program is the most reliable for reporting the correct cpu voltage? i have everest ultimate edition, intels t.a.t and ntune and they all report different voltages.
i have the bios set at 1.4 volts, everest shows 1.313 core temp 1.31 and ntune shows 1.4
Hmm... I've bumped into something strange here. Got a Pentium Dual Core E2160 the other day and have been testing it out a little. Here are the relevant specs:
E2160 @ 3.00GHz (1.30V)
Asus P5B
Scythe Ninja rev. B with 1100RPM 120mm fan
Running Orthos (Large, in-place FFTs), I get the following load temps on both cores:
CoreTemp: 74C
TAT: 72C
SpeedFan: 59C
SpeedFans CPU temp reading (Tcase, I presume) is 49C.
So, what do I do with all this? Which readout is correct? I would lean towards the SpeedFan temps, since they just seem so much more plausible. Feeling the base of the heatsink, I'd estimate it at 40C. Feeling the back of the motherboard, right behind the PSU, I'd estimate it at 50C. Also, the CPU cooler is definitely installed correctly. The setups is laying horisontally and I've tried pressing down on the heatsink and getting no difference in temps.
The IHS did seem slightly concave, but that still doesn't explain why the TAT/CoreTemp results differ so much from SpeedFan's. I thought they were supposed to match up pretty well!?
I discovered a new item to cover here...how to do the termal limit testing with Quads. I can only seem to run one instance of TAT, and test only CPU 0 and CPU 1. Is there a way to run 2 and test all 4cores?
I'm using Coretemp to determine my cpu core temperatures, and it looks like they idle in the 45-47C range (fairly high, I know). Under load they run around 58-59C. Should I start doing more on the cooling front? Here are the specs:
E2160 at 2.25ghz, vcore 1.2 volts
Standard Intel cooler, using their thermal pad
Cooler Master RC-690 using the included fans
What would be the cheapest way to drop the temp considerably? I know that the thermal pad isn't as good as thermal stuff you buy, and of course the Intel fan isn't the best. I also know that the included fans move less than 20 cfm. Last bit of info: Ambient temp is high, probably 26 or so. It gets higher than that, too.
Last question: cpu-z shows a lower vcore than I set each time (I think it was showing around 1.18 or so rather than 1.2). Is this a problem?
Sorry for all the questions, thanks for all the help!
Btw, I did run Prime on the above setup overnight, no problems for 9.5 hours.
After playing with a E2180 up to 300x10 (3Ghz, 300FSB) with a Gigabyte GA-P35-DS3L motherboard and testing it a lot, CoreTemp shows it idle at ~36ºC with C1E enabled, same as SpeedFan.
Then I tryed to OC a Q6600 with another GA-P35-DS3L motherboard, C1E enabled, but it got VERY hot, 55ºC idle, 85ºC loaded...
So I spent the day trying different configs and finished with Gigabyte's CIA2 (CPU intelligent Accelerator) set to "Sports" (7%-9% OC acording to the manual, but it does a lot more, like 10%-15%) that was the fastest stable OC with CIA2. And it run a little cooler, but still very hot.
I have it now with "Cruise" setting, (5-7%..) and now I discovered that the CPU may be cooler than what coretemp tells...
I discovered a new item to cover here...how to do the termal limit testing with Quads. I can only seem to run one instance of TAT, and test only CPU 0 and CPU 1. Is there a way to run 2 and test all 4cores?
WEll you can use coretemp for that, it monitors and measures temps on all 4 cores.
To the poster, your advice on shinetsu is spot on, i tried it and get about 3degs cooler at idle, the biggest dfference though is under load, im getting up to 8 degs lower, could, in part be due to a better reseat of my h/s but it wouldnt be that different, so thanks.
[cant get my head around the details of the guide though tbh]
I discovered a new item to cover here...how to do the termal limit testing with Quads. I can only seem to run one instance of TAT, and test only CPU 0 and CPU 1. Is there a way to run 2 and test all 4cores?
WEll you can use coretemp for that, it monitors and measures temps on all 4 cores.
To the poster, your advice on shinetsu is spot on, i tried it and get about 3degs cooler at idle, the biggest dfference though is under load, im getting up to 8 degs lower, could, in part be due to a better reseat of my h/s but it wouldnt be that different, so thanks.
[cant get my head around the details of the guide though tbh]
