Sign in with
Sign up | Sign in
Your question

Collecting Core 2 Duo CPU temps for analysis

Tags:
Last response: in CPUs
Share
September 28, 2006 12:35:09 AM

With all these apparent disparities in Core2Duo CPU temps, I would like to ask for your help in collecting some data. I know its a lot to ask, but if people post, I will correlate the data into a spread sheet and look for trends/patterns to try and identify why we may be seeing monitoring program temp disparities.:

[EDIT] due to the long load times, bandwith, and general pain for involved for folks, I am no longer requesting screen shots.

I request that you run these specific programs:
Core Temp: http://www.thecoolest.zerobrains.com/forums/viewtopic.p...
Intel Thermal Analisys Tool (TAT) http://shintai.ambition.cz/files/tat.exe
Speedfan: http://www.almico.com/speedfan.php

These programs would also be extremely helpful if you could find the time

Hardware Sensor Monitor: http://www.hmonitor.com/
Mother Board Mentor 5: http://mbm.livewiredev.com/

PCProbe if you have an ASUS mobo (comes on mobo disc)


I would like:
-temps at idle
-temps at 100% load on both cores.
-Ambient (room) temperature

To load the cores, please use CPU Burn In: http://users.bigpond.net.au/cpuburn/
set to "Disable error checking-maximum heat generation". This WILL NOT harm your CPU

You will need to run 2 instances of Burn in simultainiously to achieve 100% load on both cores


I would ask you include the following information:

(Some of this info can be found using CPUID http://www.cpuid.com/ )

Motherboard
-manufacturer
-model
-revision
BIOS
-type
-revision
- Q fan disabled or enabled -setting if enabled
CPU Heat Sink Fan/water cooling
-manufacturer
-model
CPU (Core2Duo obviously)
-version
-revision
-step
How you took the data, i.e case closed, case open, open air(mobo not in a case)
Case manufacturer /model

Any other pertinant data i.e "my case has 15 120mm fans, running dual Nvidia 7950s etc"
September 28, 2006 1:34:12 AM

E6400 @ 3200 8*400 1600fsb
Zalman 9500 CNPS
ASUS P5B BIOS 0509

TT Armor Case
2*120mm fans (intake, exhaust)
2*90mm fans (top, exhaust)
NB active cooling

@ Idle 43-44C cpu, 38C mb


@ Load 48-50C cpu, 38C mb
September 28, 2006 1:37:36 AM

Thanks for the response
Related resources
September 28, 2006 1:38:30 AM

anytime buddy, keep us posted with the spread sheet :p 
September 28, 2006 3:13:37 AM



Most of the info you need is in the sig and the pic. The bios version I'm using is F5. My heatsink is the Thermaltake Jungle 512. I have turned off fan control, thus leaving the cpu fan at max speed. My case is filled with 120mm fans, so if there is a place for a fan mount on the P180, I have a fan installed. The case was closed at the time of the readings. Also, I am using AS5. If there is any additional info you need let me know and I'll try to get it to you.
September 28, 2006 3:16:15 AM

Thanks very much for the reply.

I will post the collective data as I get it organized
September 28, 2006 3:52:33 AM

I've rounded up some readings for you, didn't use those programs cause i'm trying to keep this PC Bare minimum. I think the data is useful though being as it's similar to the other readings i saw..





Stock HSF, Antec Sonata II, w/out duct. One Huge fan in the back on High. 7600 GS (runs w/out a fan and probably adds heat to the whole senario) Ambient is usually 26-28C (80 F)
a c 133 à CPUs
September 28, 2006 4:44:24 AM

IDLE

LOAD


Please note my case was open for this....i just installed ram....
Will post again with closed results

Specs in sig
September 28, 2006 4:56:30 AM

P180B - 3 120mm in upper chamber
Active NB cooling
Zalman 9500

Ambient is roughly 22 degrees celcius.

Idle


Loaded




If you need anything else, let me know.
September 28, 2006 5:09:34 AM

thank you very much for your input
September 28, 2006 5:16:14 AM

Thanks for your input
September 28, 2006 5:17:54 AM

Thanks for your input

Quote:


Perhaps having some info about the PC's setup will help w/ the data. Case setup, Water cooling, Air cooling, ambient temp. You think?


Yes, and I did, though its at the bottom of the OP
September 28, 2006 5:47:49 AM

Quote:
Hang tight, I will get you Core Temp, Everst, and TAT for sure, on X6800 at 2.67, 2.93, and 3.2 GHz and 1067 FSB.

Also, get this stickied at the top and summarize the data.... finally, see if you cannot solicit ambient temperatures, this makes a big difference.

Jack


:lol:  :lol: 

Yup.

Typical of me...I was so intent thinking about what little things might have an effect, that I overlooked the forrest for the trees and FORGOT TO ASK FOR AMBIENTS!!!!

WHAT A DUMBARSE I AM!!!!!
a c 133 à CPUs
September 28, 2006 5:50:54 AM

Idle(closed case)

Load(closed)


Note.....I was getting 64(just a week or 2 ago)....i guess the AS5 must have set or something....If you want full fan speed just let me know...

Also good to see most people with air get the same as me....

EDIT

room temp 23-24 ish....if i trust the thermostat that controls the furnace in winter :) 
September 28, 2006 6:09:16 AM

Quote:
Just a some little improvement to make your survey better cuz I'm intrested in the data too.

CoreTemp, TAT and Everest reports all with 1-2C of each other. Ding, TM2! Those are one that's worth reporting.


Speedfan: http://www.almico.com/speedfan.php
Hardware Sensor Monitor: http://www.hmonitor.com/
Mother Board Mentor 5: http://mbm.livewiredev.com/
And the software that came with your mobo caculates through TM1 and maybe some other sensor which I don't know about.
Conclusion: Not worthy data from those above listed monitors

Other than those I think your survey would be perfect.


Wusy, you are right, but there are a couple of reasons I want the other numbers too.
One is some of the bizarre differences in mobo temps relative to CPUs temps Ive seen. The other is the guys who are seeing the same temps from PcProbe and TAT/CoreTemp....They shouldnt be

To be honest, with all the stuff Ive seen over the past couple of weeks, Im not seeing a 100% solid pattern. Everytime I think I see a trend starting to solidify, someone else comes along and posts some new numbers that errodes the theory, so, at this point I dont want to disregard a lot.

I had to force myself not to ask for ram, PSU and HDDs too. Even though these should have no apparent effect, just their physical postion in relation to the CPU HSF might have some effect.

I just dont know, so Im grasping at straws and looking for all the data (reasonable) I can link to look for a pattern.

As I get the data asembled, Ill be posting here so you and the rest of the experts can look for relations I may be missing. Im really just trying to do the secretarial work so the smart folks can look at the data.
September 28, 2006 6:35:04 AM

:oops:  sorry man. I'll redo my post and maybe play w/ those programs you listed... For SCIENCE!
September 28, 2006 6:57:31 AM

Thanks much for the input!
September 28, 2006 7:04:26 AM

Quote:
Thanks much for the input!


Couldn't get TAT to run, i seemed to get the cpu hotter when i tried to max it out w/ a virus scan while i tried to compress 700MB at the same time. Still have those old values if needed.

Andy
a c 133 à CPUs
September 28, 2006 7:12:05 AM

nice temps....what cooler? case?
September 28, 2006 7:26:32 AM

Thanks much Jack.
September 28, 2006 8:03:21 AM

Sorry haven't figured out how to add pics. Anyways, my idle temps are around 34C but at load my temps go up to 61C after about 45 seconds, after a minute temp jumps to 62C. I stop there since I get worried about damaging the cpu. By the way I'm at stock speeds and using a Zalman 7700 AlCu heat sink at full speed.

Temp monitor: Probe and Core. They both read about the same.
September 28, 2006 8:24:00 AM

Thanks Jack, I will post my numbers tomorrow since I'm currently at work. By the way, I see that you didn't use the recommended way of installing AS5 ( website instructions), but yet you got great temps. I tried to spred it the way that you did, but it didn't work for me. Temps actually went up 3 degrees. Could be my heatsink. Also have your temps dropped any since you installed it? I hear that after a week that my temps should drop 2-5c. Well I'm happy with my idle temp of 34C, just my load temp of 62C that sucks.
a c 80 à CPUs
September 28, 2006 8:38:57 AM

I assume you want any thermal management from the mobo turned off? i.e. you wants its stable temp, my fan is set to speed up at 45C and that'll drag it down even under dual prime.

Will probably have to post tomorrow. We'll see how the cheap ass mobos hold up.
September 28, 2006 9:07:58 AM

I see, I just used 100% IPA to clean the heatsink and cpu. I'll try the distilled water and acetone next time. What I don't get is why my idle temp relatively low at 34C and my load temp so high at 62C. That is a 28 degrees delta. Well EIST is enabled so at idle my freq is at 1.6ghz and 1.14v, at load freq jumps to 2.6ghz voltage 1.24v. Could it be a bad cpu? It is a ES and I could exchange it for another one, just don't want to go through the hassle if it's not that. :( 
September 28, 2006 10:28:25 AM

Quote:
I see, I just used 100% IPA to clean the heatsink and cpu. I'll try the distilled water and acetone next time. What I don't get is why my idle temp relatively low at 34C and my load temp so high at 62C. That is a 28 degrees delta. Well EIST is enabled so at idle my freq is at 1.6ghz and 1.14v, at load freq jumps to 2.6ghz voltage 1.24v. Could it be a bad cpu? It is a ES and I could exchange it for another one, just don't want to go through the hassle if it's not that. :( 

Put it under TEC+water or phasechage or SS and make good use of that 12x multiplier which isn't avaliable to retail E6700. :wink:

Well I can't see myself spending that much money to cool down a pc. Heck I reused everything but the MB from my last build. So you think I need better cooling, any chance of the cpu being bad?
a c 80 à CPUs
September 28, 2006 2:59:47 PM

Attached are idle and Hots runs (ran it for 18-20 mins, until it became stable). Not that the CPU fan is at the slow speed offer by this mobo.

Idle


And Hot


Most details are on CPU-Z.... or my Sig.

Case closed, ambient 22C, 1x120mm fan in top of case, 1x120mm fan in front of case, 1x120mm fan in PSU compartment.

Mobo, Asrock 775 dual vsta, using DDR400. Bios 1.90.

Let us know what conculsions you draw from this, if any...
September 28, 2006 5:32:23 PM

Quote:
I had to force myself not to ask for ram, PSU and HDDs too. Even though these should have no apparent effect, just their physical postion in relation to the CPU HSF might have some effect.


Depending on the rig specifics, for sure HD configuration can affect CPU and mobo temps. For example, I originally set up my video editing rig with two Raptors in RAID0 and ran the OS on that array. I ran the system a while, like a month and there was plenty of time for the thermal grease to settle, and for me to see how the rig temps behaved, etc. The case is a TT Armor and the OS array was in the ICage placed at the bottom (a HD rack that will fit 3 HDs), so a 12cm fan was blowing over them. When I was ready to add more HD capacity, I got another ICage, placed it directly above the OS rack and put two more HDs in it. That cage fan blew across the new drives and then right onto the CPU HSF. My CPU temps immediately went up 2 to 3 C at idle, mobo went up ~4C.

I could easily see the PS having a slight contribution also due to differences in efficiency and airflow. Some units pull air from the bottom, some from the back, cfms vary hugely, etc. I've seen relatively minor tweaks in case airflow have a very significant effect on temps. For example, my game box is in an Armor Jr. I initially set it up with no fan in the door and ran it a while. Eventually, I added a low-rpm fan (9dB SilenX) to the door. Granted, that's pulling air out of the box right near the GPUs but they exhaust out the back of the case anyway, so I didn't expect much change. Boy was I wrong. The mobo (DFI 3200) ran so much cooler, the NB fan dropped from 7K rpm to 3K because its temp dropped 4C. All from a minor airflow tweek. Imagine the effect of GPU!

The point of all this is that the huge variations we're seeing posted here and elsewhere are not surprising to me at all. I do not blindly assume that every person out there will do a good job of installing their HSF and its thermal grease. Etc., etc. I've seen numerous people have a very casual attitude about cooling in general. Good thread, though, just many variables to harness.
a c 80 à CPUs
September 28, 2006 5:43:46 PM

Quote:

The point of all this is that the huge variations we're seeing posted here and elsewhere are not surprising to me at all. I do not blindly assume that every person out there will do a good job of installing their HSF and its thermal grease. Etc., etc. I've seen numerous people have a very casual attitude about cooling in general. Good thread, though, just many variables to harness.


We might at least be able to spot some nodes of common temps that people should be getting close to, and provide a decent way of testing. so that we have some kind of std answer to "Is it too hot!!!..." questions. Airlfow is so so important, but that might make 2-3C difference and not the 8-10C difference of a badly installed heatsink. So spotting the 'rouge' data points would become a lot easier with a decent amount of data available.

Keep up the good work, and this is a kind of bump for the others out there who are yet to reply.
September 28, 2006 6:03:22 PM

Quote:
We might at least be able to spot some nodes of common temps that people should be getting close to, and provide a decent way of testing. so that we have some kind of std answer to "Is it too hot!!!..." questions. Airlfow is so so important, but that might make 2-3C difference and not the 8-10C difference of a badly installed heatsink. So spotting the 'rouge' data points would become a lot easier with a decent amount of data available.


Total agreement, although the 2-3C airflow effect is probably just among good airflow cases. Another example: swapping my old P4 rig from a mini-tower case to a Lian Li PC-61 and swapping the stock CPU HSF to a Zalman 7900 dropped the NB from 47C to 35, the CPU from 36 to 33 and the warmest HD from 49 to 29C. I understand that this is an extreme comparison -the mini-tower case had one 8cm exhaust fan and no intakes. The Lian Li has two 8cm intakes and two 8cm exhausts. Still - a 20C HD temp drop and 12C NB drop? Outrageous!
September 28, 2006 9:32:20 PM

Quote:


[EDTIED]Remove those useless monitors, I can see it's starting to confuss some users and they're not posting CoreTemp&TAT !!!

I will only accept data from temperture from CoreTemp&TAT and Everest, plus preferbly all the 4 CPU-Z tabs open.



Yup, I see it
September 29, 2006 1:26:42 AM

Just to keep you updated on any info that may be helpful, my ambient temp is 76 F. Also, this new pic shows more CPU-Z info and my temps at idle.

September 29, 2006 1:31:15 AM

that processor was meant to be overclocked :!: :lol: 
Why are you running at stock :?:
September 29, 2006 2:58:12 AM

Mainly because I'm lazy. :D  Also compared to my previous chip (P4 Socket 478 Prescott @ 2.8 GHz) this thing is a speed demon. I plan to OC but not yet. I mainly use this system for games and for that I will notice higher gains by upgrading my GPU. As of matter of fact, I had better 3D performance with my old system because it had a better GPU. This card was just to hold me over until the DX10 cards come out. I had to get a PCI Express card due to the fact that my old system required an AGP card and I didn't see the need to pay for a top of the line card when its successor will be released so soon. When I venture in HD video encoding, that's when I'll worry about OCing. Until then, it will do.
September 29, 2006 4:02:40 AM

RE5PECT 8)
September 29, 2006 4:09:58 AM

Thanks much for the input!
September 29, 2006 7:05:33 AM

Jack

Thanks very much for all the data youve posted. It is greatly appreciated

Thanks again to everyone else as well.
September 29, 2006 7:22:41 AM

lol, I think this was a great idea turpit.

Id join if I had a C2D I recently built a new PC month or 2 ago and bought a P5B and P D 805 only because of money. If I knew the C2D 6300 would have gone 40$ within a couple weeks or so and I would have to buy a Zalman 9500 for this dam hot 805 I would most definetly gone for the 6300 instead. Eventually I will. =\

But anyways when will you be posting your spreadsheet or whatever Id really like to see it also.


Thanks Storm
September 29, 2006 7:33:25 PM

Firstly Im not overclocking. And secondly there's more to it all that Im not gonna go In depth about it, simply put I shoulda went ahead and got a 6300 instead of a 805.

Just to throw out there im not a person that has praised the 805 at all.
As I said I just got it because I figured Id save 100+$.

And coming from a Celeron 2.0ghz I figured it would blow away it which all and all it has not overclocked any.

So I mainly got it for the price and upgradability in the future, hopefully the P5B regular will support quad core and someday Ill throw one in.


Storm~
September 30, 2006 12:09:55 PM

EIST enabled:
temp at idle 34 C
temp at full load 40 C
ambient temp 26 C





my gear:
intel core 2 duo e6300 @ stock speed & heatsink fan
asrock 775dual-vsta
2x corsair vs1gbddr400c3
his ati radeon x1600pro iceq II ddr2 256mb agp
western digital caviar se 8mb cache SATA II 120gb (jumpered to SATA I)
seagate barracude 7200.7 8mb cache SATA I 120gb
benq dw1670
leadtek winfast tv2000xp expert
lg flatron l1530s
simbadda 500watt psu with 2x 8cm fans (1 rear exhaust, 1 bottom intake)
thermaltake armor jr. casis with 2x 12cm fans (1 rear exhaust, 1 front intake)
thermaltake icage 1x 12cm fans (1 front intake)
October 1, 2006 4:02:09 AM

I was runnning your tests, and I couldn't get my CPU load to 100% for some strange reason. I ran Intel's Processor ID Utility and got this screen:



Do I win?

I will update this post when I can get some real data. :D 
October 1, 2006 4:47:57 AM

Quote:
did you run prime95 on both cores? If you didn't here's a guide on how to set up your prime95 to stress two cores dual Prime95 for stability testing


No I didn't.

I was simply trying to run Turpit's burn-in test, and I ran it for 30 mins, and the CPU loads were bouncing between 55% and 75%, but never hit 100%.

My temps never went over 45/43C, but I don't think my CPU was at full load. At idle, my CPU sits at 36/34C.

I can try other programs that will max out my CPU load, and try to get a more stable temp, and maybe get a screenshot of it.

Thanks for the link. I never did use Prime95 for anything. Never thought I ever needed it. LOL
October 1, 2006 4:52:31 AM

post your results after you get temp @ load
a c 80 à CPUs
October 1, 2006 11:50:13 AM

Quote:
did you run prime95 on both cores? If you didn't here's a guide on how to set up your prime95 to stress two cores dual Prime95 for stability testing


No I didn't.

I was simply trying to run Turpit's burn-in test, and I ran it for 30 mins, and the CPU loads were bouncing between 55% and 75%, but never hit 100%.

My temps never went over 45/43C, but I don't think my CPU was at full load. At idle, my CPU sits at 36/34C.

I can try other programs that will max out my CPU load, and try to get a more stable temp, and maybe get a screenshot of it.

Thanks for the link. I never did use Prime95 for anything. Never thought I ever needed it. LOL

Odd result....

The prime tests may give a differebt temp than the burn in test even though they are at 100% load. I get 2-3 C hotter on prime than I do with the burn in test, but if everyone uses the burn in test then the results will be more comparable.
October 1, 2006 8:12:01 PM

i used prime95 to test stability after i oced to 8*400
in my opinion prime95 is the best 8)
a c 80 à CPUs
October 1, 2006 9:10:52 PM

Quote:
i used prime95 to test stability after i oced to 8*400
in my opinion prime95 is the best 8)


I also use '95 for stability, the tests here are to get a consistent and knonw processor load from many different people, and a piece of software with only 1 setting seems like a better way of doing that, than prime which will test using as much RAM as you want it to, which may or may not affect how much heat it generates.

Like I said, I get more heat off prime, but the other software would give a consistent load across everyone that used it.
!