Penryn Temperature Software Discussion [update]

[Tempered81]

http://www.xtremesystems.org/f...howthread.php?t=179044


It also explains the problem with the 45nm chips and reading their DTS.

It appears that both situations are correct. There are occasions where 45nm intel chip's DTS sensors can be permanently "stuck" and read incorrect data at idle and load, and there is no fixing this issue. All other chips that do not exhibit this flaw are being read incorrectly by current software, for the most part. If your chip does not have the "stuck" DTS, then it can be read correctly at idle and load now with RealTemp, compared to incorrect data reported by current temperature software such as coretemp/everest/etc.

download realtemp 2.1 here:

http://www.fileden.com/files/2...3/1794507/RealTemp.zip


download x64 bit version here:
http://www.fileden.com/files/2...1794507/RealTemp64.zip

 

[Tempered81]

Another interesting point they seemed to have discovered is that TJmax is not 105, but rather somewhere between 92.4C and 95C. Coretemp was reporting it wrong. This means that when coretemp was reading 70C, RealTemp is going to read 60C. That seems kinda strange, think I'll wait until more people have tried this out to make a steadfast conclusion. Would be nice to find out that the whole time you thought you were running 65C load, it was really 55C load. hahahah

edit: I'm talking about e8400's.

 

[aigomorla]

nice find jared.

Must of missed that thread last night.

:T


But definitely very interesting read.

 

[nullpointerus]

does this impact earlier chips?

 

[CurseTheSky]

Sounds great. I'm glad to see people doing research on this. Seems the general concensus is Core Temp is reading about 10C high at full load.

 

[imported_wonderwrench]

I do not have a E8xxx but have a E6400 L2. He also used this CPU as a comparison.
It seems Real Temp thinks my CPU has a Tjmax of 85C and not 100C.
Everything I have read says the E6400 has a Tjmax of 100C.
Now E6400's in general seem to run hotter than say a E6300 at the same clock speed and voltage. Both CPU's have a TJmax of 100C as far as I can tell.
This has always bugged me. Something is not being accounted for.

I have checked and calibrated speedfan's reported Tcase and Tjunction temps using Computronix Guide here. http://www.ocforums.com/showthread.php?t=543522
Now this guide could be all wrong but it seems to make logical sense and the results make sense. Tcase calibration would seem to be almost infallible IMO, at least when the CPU is at idle. Tjunction temps are a gray area do to Intel not releasing compete info.

Now I believe I can trust the Tcase temp reported by Speedfan at idle for sure because of the method of calibration in Computronix guide. Load temps are more questionable because there is no easy way to check accuracy. I'm not using a Nvidia based mother board which is known to report Tcase temp in a non linear way. So I'm going to just trust Tcase temps reported by Speedfan to be accurate. I also believe Tcase to Tjunction delta's on E6400's are 15C +-3C as the guide claims. Now Real Temp set to ++ as Unclewebb recommends on a E6400 reports Tjunction temps way to close to speedfans reported Tcase temp at idle IMO. Now Unclewebb claims E6400 CPU's under report Tjunction temps at idle but using Real Temp set as he specifies for a E6400 reports lower Tjunction temps than any other program. Speedfan (calibrated), Coretemp, HWmonitor etc. I just feel something is wrong here.

Looking at load temps Real Temp reports Tjunction temps lower than Speedfan reports Tcase temp. This is not possible so one or the other is wrong. I still wonder what the true Tjmax temp is on a E6400 really is? Load temps sure look nice using Real Temp compared to other programs but something just does not feel right.

Some screen shots
Idle
http://i156.photobucket.com/al...erwrench/Misc/Idle.png
Load
http://i156.photobucket.com/al...wrench/Misc/load-1.png


In the end I say not enough is known about Intel's Core2 CPU's to be able to trust temps reported by any method. I think Unclewebb is on to something but too many pieces of the puzzle are still missing.

 

[Tempered81]

Originally posted by: wonderwrench
I do not have a E8xxx but have a E6400 L2. He also used this CPU as a comparison.
It seems Real Temp thinks my CPU has a Tjmax of 85C and not 100C.
Everything I have read says the E6400 has a Tjmax of 100C.
Now E6400's in general seem to run hotter than say a E6300 at the same clock speed and voltage. Both CPU's have a TJmax of 100C as far as I can tell.
This has always bugged me. Something is not being accounted for.

I have checked and calibrated speedfan's reported Tcase and Tjunction temps using Computronix Guide here. http://www.ocforums.com/showthread.php?t=543522
Now this guide could be all wrong but it seems to make logical sense and the results make sense. Tcase calibration would seem to be almost infallible IMO, at least when the CPU is at idle. Tjunction temps are a gray area do to Intel not releasing compete info.

Now I believe I can trust the Tcase temp reported by Speedfan at idle for sure because of the method of calibration in Computronix guide. Load temps are more questionable because there is no easy way to check accuracy. I'm not using a Nvidia based mother board which is known to report Tcase temp in a non linear way. So I'm going to just trust Tcase temps reported by Speedfan to be accurate. I also believe Tcase to Tjunction delta's on E6400's are 15C +-3C as the guide claims. Now Real Temp set to ++ as Unclewebb recommends on a E6400 reports Tjunction temps way to close to speedfans reported Tcase temp at idle IMO. Now Unclewebb claims E6400 CPU's under report Tjunction temps at idle but using Real Temp set as he specifies for a E6400 reports lower Tjunction temps than any other program. Speedfan (calibrated), Coretemp, HWmonitor etc. I just feel something is wrong here.

Looking at load temps Real Temp reports Tjunction temps lower than Speedfan reports Tcase temp. This is not possible so one or the other is wrong. I still wonder what the true Tjmax temp is on a E6400 really is? Load temps sure look nice using Real Temp compared to other programs but something just does not feel right.

Some screen shots
Idle
http://i156.photobucket.com/al...erwrench/Misc/Idle.png
Load
http://i156.photobucket.com/al...wrench/Misc/load-1.png


In the end I say not enough is known about Intel's Core2 CPU's to be able to trust temps reported by any method. I think Unclewebb is on to something but too many pieces of the puzzle are still missing.

I posted what you wrote for the Realtemp author "unclewebb" to read. This is his reply:

jaredpace: CPU diode based temps reported by motherboards are rarely 100% right. You can't assume anything in this and can't just assume that what SpeedFan reports must be right for the CPU or any other piece of software. I know personally my CPU diode based temp on my P5B can change by 10C depending on whether I boot up or do a resume from Stand By. I definitely don't trust anything that sensor tells me because it is not accurate. Try dropping your CPU down to a low wattage setting like Slay0r did and report what it says.

 

[aussiestilgar]

Installs and runs fine on Vista HP 32. Report my E8400 temps 10c lower than Coretemp, and to my relief shows that my core sensors are not stuck at all. Temps seem more in line of whats expected and I like the ability to skew the idle temperatures albeit if it is guesswork.

Cheers.

 

[Big Lar]

Installed it early this morning and it's working fine. Nice to know what load temps actually are on this thing

Larry

 

[Markfw]

Good find jaredpace, I see a lot of people here are very happy with you !

 

[kentsfield]

It`s not working with Vista x64 SP1. " Make sure the driver you are using is 64 bit bla bla bla"

 

[imported_wonderwrench]

Originally posted by: jaredpace


I posted what you wrote for the Realtemp author "unclewebb" to read. This is his reply:

jaredpace: CPU diode based temps reported by motherboards are rarely 100% right. You can't assume anything in this and can't just assume that what SpeedFan reports must be right for the CPU or any other piece of software. I know personally my CPU diode based temp on my P5B can change by 10C depending on whether I boot up or do a resume from Stand By. I definitely don't trust anything that sensor tells me because it is not accurate. Try dropping your CPU down to a low wattage setting like Slay0r did and report what it says.

I agree Tcase temps are usually off by a mile. He must have just blew by the fact that my Tcase temp is not bogus. At idle its right on and I assume its close under load but have no way to prove it. The Tcase temp problem he is reporting is a bios bug, the sensor it self is fine. It's the fact that Tcase temps are reported by the bios that causes most of the problems with Tcase temp readings. Tjunction temps are read directly from the CPU and as long as Intel did not drop the ball, all thats required to accurately read and report Tjunction temps is software smart enough to do it. Trouble is Intel does not want anyone to know the real story.

I retested my calibration of Speedfan using Computronix method, which by the way is very similar to Unclewebb's method. Both lower the CPU down to 1.6 Ghz at low voltage.
Here is a screen shot taken at idle, all fans at 100% and case covers off. Note digital thermometer measuring air temp in front of CPU cooler intake fan. Of course this was not totally live as seen in the shot. I took a picture of the displayed temp, opened the result on my desktop then hit print screen. http://i156.photobucket.com/al...nch/Misc/Idle_shot.jpg


I still think something is up with the temp reporting using Real Temp at least on E6400's.
It is possible that Tcase to Tjunction delta could be only 1C-3C but load temps just do not look right to me. Maybe its just me? Not saying Unclewebb would do such a thing but dropping the Tjmax temp down to make the load temps look nice would cause idle temps to be reported below ambient temp. So applying offsets to temps in the idle range would cover this up. This would be no different than everyone else upping the Tjmax to keep idle temps above ambient temp. What is right who knows, I doubt Unclewebb would go through all the work he has done just for the fun of it. If I work up the balls I'll try to reproduce his results on my E6400. What sucks is I just mounted the heat sink not long ago and really do not want to take it off again.

 

[Tempered81]

if you click there link in the top post and goto the XS post he made, someone has posted a graph of temperatures in coretemp and realtemp and given a statistical explanation for it. I would recommend you check that out.

 

[BlueWeasel]

From an idle state, running Test Sensors indicates that my Core #0 is bad. It reads 10C-12C +/- higher than Core #1 when loaded. Does that mean I just need to ignore Core 0?

With an E8400 running at 3.9Ghz at 1.3v with a Scthye Ninja and 120mm fan, Core #0 hits 70C and Core #1 never goes above 60C. Which one to trust?

 

[n7]

No x64 support i understand?

Sounds like Everest will continue to be my measurement tool of choice.

 

[M1A]

I downloaded ia32 and installed but I do not see a real temp exe to run the program?

 

[VirtualLarry]

"Using this knowledge and some more measurements, I came up with some simple correction factors to help compensate for errors in the DTS at idle. Using a setting in Real Temp of (--) for my E8400 and using a setting of (++) for my E6400 now gives me very accurate temperatures for both processors from 10C to 100C. Typically within 1C or 2C of what an IR thermometer shows near idle when pointed at the bare core and exactly equal higher up. The formulas were designed to bring the above two processors a lot more in line with each other which they should be given the identical core voltage. Power or heat generation is proportional to the square of voltage so that is the most important factor to equalize when making comparisons like this."

He's making two mistakes. One is that he is reporting temps normalized to an empirically-determined Tcase value, not Tjunction.

The second is that temps for chips clocked the same, and at the same voltage, will not be the same, because of differences in the silicon. (Leakage, among other things) Attempting to normalize this is wrong, IMHO. Even moreso that he calibrated his program using only two sample chips that he had.

This program doesn't sound that useful to me, because I doubt its correctness.

 

[VirtualLarry]

From RGE's post in that thread:
"Only issue I have is, I calibrated my cpu temp using lasertemp (+/- 1C accurate and checked against very accurate temp probe) so cpu temp in speedfan always reads within 0-2 C of IR to cpu from 45-95C.

If Tjmax on my E8400 is 95, then there is no gradient from core to tcase on full orthos load, as my calibrated cpu temp is now same temp as core. And I can show for example, that when speedfan cpu shows temp of 53, IR to back of cpu will be same.

Unclewebb, you could easily calibrate your cpu, if have not already, and test same, would be interested in what you get. When I reach 105C core, my tcase measured by IR was 94.8, so I got the same result as you. The only issue is, does any gradient exist between core and tcase at that time. Your control chip, one did not. On my E6850, a gradient of 5-7C existed at all times, and on my E8400 a gradient of 10C existed at all times."

I think he's asking why RealTemp returns temps that are akin to Tcase temps, instead of Tjunction, and that explains why UncleWebb thinks that Tjmax is 95C instead of 105C. So it sounds like 105C is actually correct, if there is a 10C delta, and he measured 95C at Tcase with the IR thermometer when throttling occurred.

Edit: UncleWebb's reply:
"rge: I did all my IR testing with the CPU at idle. Even when it was over 100C it was still only at idle. Temperature gradients between TCase and core temp exist but I believe that when at idle, the difference between the true core temperature reading and what you can measure on the top side of the IHS is so small that it is within the accuracy of the test equipment espicially at lower temperatures where I was testing."

He's saying that there is NO delta between Tcase and Tjunction while at idle. This I cannot believe, because it depends on the device physics. I have a hard time believing that this delta scales up as temps increase.

In all the cases that I've seen, this delta is relatively constant, and is definately non-zero, at all temps. That's what RGE was saying too, I think.

So from what I've seen, the premise behind some of this program's calibration seems to be off.

 

[Tempered81]

hrm. I feel dumbfounded after reading larry's & webb's posts. Well everyone thinks coretemp is wrong on e8400's, you think realtemp could be wrong on 8400's as well? I've seen a lot of threads pointing out that 105C reported by coretemp for Tjunction max on e8400's is incorrect. Larry, what did you make of ixtap's Statistical testing in the thread, (it's the post with the graphs)?

 

[ixtapalapaquetl]

Originally posted by: jaredpace Larry, what did you make of ixtap's Statistical testing in the thread, (it's the post with the graphs)?

Hey that's me! But please, call me ix - ixtap is what my mother calls me.

 

[LOUISSSSS]

just installed RealTemp 2.0 on my XP 32-bit w/ Q6600,

reports similar temps as coretemp:

RT: 49,41,42,46 (Core Temperatures) TjMax = 95C
CT: 54,46,47,51 (CPU #0 Temperature Readings) TjMax = 100C
about 5 degrees delta between the two programs, with RT<CT.

side note, i think that CT 0.96 is reporting temps that are "Distance to TJ Max," CT's readings seems more in line with RT's Distance to TJ Max: 48,54,53,50 (if u rearrange the core readings they are pretty similar.

SF: 51,51,51,51 <~ i don't find speedfan 4.33 accurate on temps at all since it shows all my core temps to be within 1C range at all times and never deviates from there but i'm sure that my tuniq mounting is uneven on one side...just too lazy to fix it.

can someone explain to me what TEST SENSORS does?

 

[LOUISSSSS]

RT uses about 3952k of ram
CT uses about 7756K of ram

i like small programs =]

 

[Anubis]

Originally posted by: n7
No x64 support i understand?

Sounds like Everest will continue to be my measurement tool of choice.

that would explain why it didnt work at all for me

 

[Germonicus]

It seemed to support Vista 64 for me.....just copied the .dll and .sys files to the appropriate directorys in windows and voila.

Sadly it looks like my CPU has stuck sensors as it reports 0 movement on either core of my E8400

Now to the BIG question,if the temp reporting is screwed up for my CPU then HOW do I know that my OC isn't running much hotter than I was prepared to accept?
Also,is this a fault that needs to be RMA'd (chip is only 3 weeks old)?

 

[PClark99]

gotta try this out later.