Doubtful temperature readings on M0 stepping E2180

[mrQQ]

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..

 

[CompuTronix]

mrQQ, it's difficult for anyone to help you if you don't list your system spec's. There's a format for this in the Temp Guide Sticky above, which is shown as Results and Variables in the Introduction, and again in Section 10. The system information you've given doesn't leave much to work with. If you do some research, you'll find your answers in the Guide. Here are some hints:

CPU = Tcase
Core = Tjunction

From the Guide:

Section 6: Scale

E2xx0: Tcase Max 73c, M0 Stepping, Tjunction Max 100c, Vcore Default 1.312, TDP 65w, Delta 10c

-Tcase/Tjunction-
--70--/--80--80-- Hot
--65--/--75--75-- Warm
--60--/--70--70-- Safe
--25--/--35--35-- Cool

Section 8: Tools

Core Temp 0.95.4 (Beta): http://www.alcpu.com/CoreTemp

Note 4: The software utility (TAT) Thermal Analysis Tool at maximum settings will simulate 100% Thermal Load, which would equate to Prime95 at 114% Workload and ~5c hotter. This provides the most extreme testing available for CPU and system cooling efficiency. Since TAT is coded to measure Notebook temps, it identifies a C2D as Pentium M. As Notebooks have no Integrated Heat Spreader, thermal scaling differs from desktops CPU`s, so TAT indicates ~ 2c lower, and depending on Variables, temps may be Offset by more than 15c. It is therefore recommended that TAT be used for extreme thermal testing only, and temps regarded as unreliable.

Comp

 

[mrQQ]

Oooooh, i was thinking that throttling kicks in when any of core's reach 73..

for me @ 3.0ghz and stock voltage i get

--48--/--60--59-- (that is with +15 added to coretemp readings)

so i guess i'm pretty cool there, and can go for much higher voltage, correct?

 

[CompuTronix]

One more time... it's difficult for anyone to help you if you don't list your system spec's. There's a format for this in the Temp Guide Sticky above, which is shown as Results and Variables in the Introduction, and again in Section 10.

And without knowing Ambient and Vcore... ???

Read Section 12.

Comp

 

[cnumartyr]

One more time... it's difficult for anyone to help you if you don't list your system spec's. There's a format for this in the Temp Guide Sticky above, which is shown as Results and Variables in the Introduction, and again in Section 10.

And without knowing Ambient and Vcore... ???

Read Section 12.

Comp

Saint

 

[CompuTronix]

You're evil...

 

[cnumartyr]

You're evil...

lol.. I'm relapping and I'm bored man cmon!

Ultra Kaze's came in today, 120x38. 134 CFM and 3000 RPMs.

Might have some fun with a Push/Pull after I get things back.

 

[CompuTronix]

That's some serious airflow.

 

[mrQQ]

Ok, sorry once again hope i get this correct:

Tcase = 37c Idle, 58c Load (TAT 100% load)
Tjunction = 47c Idle, 70c Load

Ambient = 22c
Chipset = P35
CPU = E2180
CPU Cooler= Zalman 9700 NT
Frequency = 3.0 Ghz
Load = TAT 100% load 5 minutes
Motherboard = Asus P5K/Wi-Fi Premium
Stepping = M0
Vcore = Auto (results in ~1.305 on load)

did i get it all correctly now?

 

[CompuTronix]

No. DO NOT use TAT for Load testing. Read Section 8 - Note 4, then read Section 9 - Part 5 which states "Load Test = Prime95 - Small FFT's" - 10 Munites. Heat saturation requires 7 to 8 full minutes, which is why the Guide specifies 10 minutes. Please follow Part 5 to the letter.

Comp

 

[mrQQ]

uuugh, seems its hard for me to get things right

ok, Prime95, 10 minutes load:

Tcase = 37c Idle, 51c Load
Tjunction = 47c Idle, 62c Load

interesting thing to note, that it doesnt matter if my CPU cooler is running at 100%, or at 50%, Load temps are still same.. and the heatsink is cool to touch.

 

[CompuTronix]

Prime95 - Small FFT's, not Blend.

 

[mrQQ]

It was Small FFT's

 

[CompuTronix]

Why is your Vcore in Auto?

 

[mrQQ]

Because it runs stable in 3.0, and i enabled EIST, which automatically drops multiplier AND vcore when not needed, thus power savings..

 

[CompuTronix]

You realize that Auto applies excessively high Vcore during Load conditions, which results in higher temperatures. Since your temperatures are good in Auto at 3.0, you could run 3.2Ghz in Manual, and not exceed 60c Tcase and 70c Tjunction. It would simply require some additional stability testing.

Comp

 

[mrQQ]

well, my CPU only runs 3.2 with 1.45v stable, any less and it crashes. And auto applies 1.305 in Load, which is less than stock voltage for this CPU, and 1.224 on idle, which is really low, so considering that most of the time i'm in idle mode, i think i'm saving much more..

 

[mrQQ]

sometimes it drops to as low as 1.184 on idle.. and if i set manual mode, then it wont change voltage automatically, and i need at least 1.3 to be stable at 3.0..

 

[CompuTronix]

As you prefer, however 1.45 is not too high, and is actually the "sweetspot" for an optimal overclock, provided safe temperatures are maintained.

 

[mrQQ]

wouldnt lifetime of cpu be quite shortened when running 1.45 vs running 1.3?

also, if i set manual vcore, then i dont get advantage of it dropping to lower during idle, and 200mhz difference isnt such a big deal, no?

 

[CompuTronix]

If CPU longevity was an issue at 1.45, then I wouldn't be running at that Vcore myself. Additionally, Intel recently changed their Vcore Max specifications to 1.5. If you don't have any use for the additional overclock, and you prefer to run at 3.0Ghz with Auto Vcore, that's certainly you perogative, and it's not my place to debate that with you, so suit yourself.

 

[mrQQ]

I was just asking perhaps i missed something important, and it's always good to know other opinions

So, anyway, how do my temps look?

 

[CompuTronix]

Excellent! What were your Idle Test Results?

 

[mrQQ]

I didn't do it.. it appeared very complex for me *ashamed*..

 

[CompuTronix]

Well STUPID ME !!! And here I am assuming that you were presenting valid test results! You would think that I would know better by now! Did you even TRY to perform ANY of the Calibrations? I retract my assumptions. I have no bloody clue as to whether or not your temperatures are accurate... do you? Let me consult with my Crystal Ball and Ouija Board, and I'll get right back to you... :lol: