Last message on previous page:
my e6300 is running at 2.2ghz on stock and my tems change drasticly when under load..
speed fan;
idle -CPU =40c
-Core 0 =32c
load -CPU = 69c
-Core 0 =53c
wtf...?
So as not to repost the same things, I wanted to link my question. Your help is aprreciated
http://forumz.tomshardware.com/har [...] 24#1607624
You can measure your tjunction either with TAT or my new favorite RMClock which has much more features including logging over TAT.
OP: can you please add the link to RMClock in the first thread?
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.
you 've gone to alot of effort here
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
Hi,
On my set-up I have the opposite of what's expected:
TAT and CoreTemp Report the following Temperatures when idle:
Core0 = 32C
Core1 = 30C
Under Stress this goes to around 50C.
However, PCProbe & SpeedFan report a CPU Temperature of 53C when idle and 71C when stressed.
What gives?!
Hi,
On my set-up I have the opposite of what's expected:
TAT and CoreTemp Report the following Temperatures when idle:
Core0 = 32C
Core1 = 30C
Under Stress this goes to around 52C.
However, PCProbe & SpeedFan report a CPU Temperature of 53C when idle and 73C when stressed.
What gives?!
hahaha
hahaha
Hello out there, need some help.
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.
Trust your temps?
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...
Check this out
What software can I trust for Core2Quad?
And if ~56ºC idle is the right one, it seems very hot for me, is it normal?
Any advice?
Thanks!
| TexasVol wrote : 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]
| TexasVol wrote : 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 calibrated Tcase according to the guide, and it was about 4C too high. Nice Guide! I will post overclocked info later.
Tcase = 26c idle, 43c load
Tjunction = 34c idle, 51c load hottest core #3
Ambient = 25C
Chipset = nForce 680i SLI
CPU = Q6600
CPU Cooler = TR Ultra 120 Extreme
Frequency = 2.4 Ghz (Stock)
Load = Prime95 - Small FFT's - 10 min
Motherboard = EVGA 680i SLI
Stepping = G0
Vcore = 1.25
Heat Score
(A) Ambient = 3
(B) CPU Cooler = 1
(C) Computer Case Cooling = 1
(D) Frequency = 1
(E) Graphics Cooling = 2
(F) Hard Drives = 2
(G) Vcore = 1
Total Score = 11 (Safe)
Hi ... thanks for that great thread..
But I have a weird thing !!!
My system is Gigabyte 945P-S3 rev3.3-C2D E6420-2GB 667 ram-Geforce 7900GT PCIexp
I have the Tcase is HIGHER than Tjunction1 and T junction2 !!!
being for ex. 43c/32c---33c !!! is this normal at idel ? my system is not overclocked or something ...everything is at stock ...
And during gaming ..is hitting the 50c in any parameter is bad or not really bad ?
by the way...I didn't apply any thermal paste during the CPU installation and I'm using the stock intel fan cooler..according to the buyer , It doesn't need thermal paste cause the cooler contains it at the buttom !! Is this true or I should use a thermal grease ??? thanks in advance...
I dont know if im blind, but i have ben looking and searching without sucsess.
I have a Q6600. the "problem" is that one core is way hotter then the others. 10+C in load like 56 44 44 45. When idle its about 5C over. Is this normal? 56 44 thats nearly 30% hotter on one core.
What have I done.. replaced thermal paste (using thermalrights extreme 120 cooler), tried lots of different themral tools.
10c!! that one core messes everything. cant clock as i want if that one is 10c or ever higer! .. anyone know about this?
I am running my q6700 at stock speed, and when doing this torture test in prime95, my cores goes up to about 66-67C.(34-35 idle) Is this a safe temp? btw I am using the Thermaltake MaxOrb CL-P0369 cooler, and could that be the cause of the problem? Also, I probably wont stress the cpu as much as in prime, so I guess i wont reach those temps outside that prog, and does that mean the temp is safe?
All help appreciated.
newbie here... something is wrong with my readings and I'm not sure how to offset them... maybe one of you can help.
I have an E6300 (L2) on a GA-965P-S3, and at idle speedfan reports:
Temp1: 43c
Temp2: 40c
Temp3: -1c
Core0: 33c
Core1: 33c
Core: 63 (I'm guessing this is the GPU, which is a 8800GTS 320mb)
Ambient: 0c
This is all while OC @ 2331 MHz and core voltage at 1.2V, but it's pretty much the same at idle stock...
ok. i have a question - are you guys really really sure that M0 stepping of E2XX's have to add 15c to CoreTemp's readings? because this makes no sense - i'm running TAT at 100%, it's saying 59/54, if i add 15c that would be 74/69, it should be throttled already, but it's not, and my CPU heatsink isn't even warm to touch! something is terribly wrong here..
i'm terribly sorry - will do
Very helpful guide. Now I can actually trust my SpeedFan temp readings.
E6750 @ 400x8 (3.2Ghz), 1.35V
Antec 900 Case
CPU Cooler = AC Freezer 7 Pro (set to ramp up to 100% as temp approaches 40C)
Ambient = 22
Tcase Idle = 25, Load = 40
Tjunction Idle = 32, Load = 50
Love the temps, now I can finally sleep fine w/o worry...
I wonder what integrated graphics will go under. What heat score will it get?
Hi there. i´m sorry but still confuse with temps. i don´t know wich software to trust, wich is my limit temps, etc.
i´m doing some overclock and what i want to know is wich are my safe temps and where should i stop pushing.
I have an intel e2160 Dual core 1.8Ghz CPU (Stepping 2 Revision L2) OC at 3.0 Ghz
According to your article my temps should be(1):
75-75 - Hot
70-70 - Warm
65-65 - Safe
40-40 - Cool
Ok. Temps Reports are:
CoreTemp SpeedFan Intel TAT
Idle 45-42 27-27 47-45
Load TAT 71-70 54-54 70-68
So wich should i compare to (1)?
lots of thanks
From the looks....I say trust Core Temp....Speedfan is too low unless your using aftermarket cooling
"If temperatures don't meet Parameters, then SpeedFan 4.33 can configure Offsets to correct for Tcase (CPU or Temp x) and Tjunction (Core x). From the Readings tab, click on the Configure button, then click on the Advanced tab. Next, click on the Chip field directly under the tabs, then use SpeedFan's Help and HOW-TO icon included in the installation Program Group."
i dont understand that.. i have a Q6600 with G0 stepping and i dont see any how to.. could someone maybe give me a step by step on how to do it please
The guide has got nice and short lately ![]()
In case you need it, google's cache has the remainder of the guide here.
This guide is exactly what I needed to understand the relationship between Tjunction and Tcase on our C2D, C2Q processors. However the guide is now missing sections 6 through 16: "SERVER EDIT MALFUNCTION ... REMAINING CONTENTS TEMPOARILY UNAVAILABLE ... PLEASE TRY LATER." Is this archived anywhere that I can retrieve it? Thanks!
Look at the link in my previous post above, that is a google cached version from 6 Dec 07, so it may not be totally up to date. It should be good enough though.
The google cache page has been "updated" to the broken guide now, damn you google!
^ You should mention the problem with the E8xxx temperature probe, which causes incorrect readings in CoreTemp and SpeedFan.
Edit: Saw your post later
![]()
| CompuTronix wrote : Shadow703793, I'm presently working on including this problem in Section 15 (Troubleshooting) of the Core 2 Quad and Duo Temperature Guide: http://www.tomshardware.com/forum/ [...] ture-guide The faulty temperature sensor issue is not confined to the E8000 series processors, but is also common to the Q9000, and QX9000 series processors. |
http://www.tomshardware.com/forum/ [...] 28_40.html

@CompuTronix: You should do a temp guide on the different chip sets (ie, P35,X38,680i,etc) also if possible.

i'm sort of confused. from what i understand, there is a possibility my temp readings are inaccurate.
so i grabed a thermal gun and scanned the ambient temp while the system is off and it read 18C. with that, should the Tcase automatically be 19-22C? and based on that temp, should i offset speedfan to be +5 of tcase?
thnx
"Core 2 Quad and Duo Temperature Guide"
Huh? To someone with a degree in Physics?
I finished Uni and came out with an Electronics Eng. This "Guide" as you put it....wow!! what typical Overclocker will get to grips with all of this jargon?!
We dont mind being educated....but some of these "guides" were written by people with no teaching or writting skills whatsoever.
^ Agreed with Comp.
@Steve_long: This guide is meant to provide very detailed information. If you can't handle the details you should look else where.
Random thought:
Comp, what's your take on RealTemp?
http://www.xtremesystems.org/forum [...] p?t=179044
PM me with your thoughts.

| Steve_long wrote : "Core 2 Quad and Duo Temperature Guide"
|
| Steve_long wrote : "Core 2 Quad and Duo Temperature Guide"
|
I agree 100%
I think most of the people saying "great guide" are just trying to sound smart. This is rubbish, Everest is WAY better than speedfan.
I mean how in the world am I supposed to calibrate this.
I will continue to completely ignore Speedfan and stick with CoreTemp.
There are 1026 identified and unidentified users. To see the list of identified users, Click here.
