"Core Temp" - A program for accurate K8 temperature monitoring

[The Coolest]

Question:

How low temps would this sensor be able to read from a conroe? I will be phase-cooling one soon to very low (far sub zero) temps.

-Stigma

I don't know. The lowest reading possible in theory is TCaseMax - 127C. But the lowest I've seen this far was 4C even at sub zero cooling.

[MikalCarbine]

Just got my rig back up, and I have some cool results



This is a naked 4200+ X2
There is a 4C difference in temps between the two cores at idle, and MBM5 seems to report cooler core temp (or just Core2's temp)

At load though, Core1 spikes all the way up to 40C! I don't think that it is a bad paste job, because I have had many problems overclocking this chip. No matter what I would do Core1 would ALWAYS fail really fast, while the other core remained rock solid, most likely due to the temperature difference. Has anyone else seen this?

Hopefully I can find a fix for this and get some nice clocks :P

BTW, does anyone wana test a CPU in a 4 phase mobo compared to a 3 phase and see what the "real" temp difference is?

[Kalway]

C0 core... that mean it'll work on Newcastles?

[ahmadmz]

my core 1 is 38C and core 2 is 40C .. im using the stock Opteron 165 HSF and AS5 .. why 1 core is like 2-3 C higher than the other ?? And i heard Opteron hsf are good.. i also just removed stock pad and applied AS5 .. still hot!! Is there something i can do/check to make it better or this is normal?

[lsevald]

I'm getting -26°C with a stock MachII unit (Core Temp v0.9.0.86), which seems a bit optimistic...but otherwise great app. Works fine on my other computers (air and wc cooling)

[MikalCarbine]

Following up with my post, if I prime just Core1, its temp stays near idle/low, but after I kick in Core2 (even if Core1 isn't being primed) it spikes up to 40 instantly

[The Coolest]

C0 core... that mean it'll work on Newcastles?

Yes. Newcastles are CG

[CanadianTSi]

After some testing I think that core 1 may be located on the left side of the chip, that is closest to the back side of the case...

[MikalCarbine]

Just redid my paste job and it didn't help Core1's temps at all :S

[dinos22]

I'm getting -26°C with a stock MachII unit (Core Temp v0.9.0.86), which seems a bit optimistic...but otherwise great app. Works fine on my other computers (air and wc cooling)

something's wrong there...........that's a massive difference........i'd say your smartguardian temp is the correct one there rather than the program unfortunately..........so around -1C...............hmmmmmmm can anyone else try subzero temps

[lross78550]

I did try the program with my mach II this post was on page 5 but since i didnt have any screen shots i guess no one read it.
my temps are listed in the text below
Lee

I noticed a big difference in the temps on my a8n32-sli and fx-57 the asus temp program and speedfan both report my cpu temp as being -12c at idle in windows. your program reports the core temp as being -28c (im running a prometeia mach II) I hope you program is right as being i like your numbers better I really think you should have the title of your thread changed now that you have it working on processors other than the Rev F chips. and i think you should have it stickied on one of the forum topics, finally there is a way for people with fairly current hardware to get accurate cpu temps. As soon as my jinuv2000 ss phase shows up and i build my fx-62 system i will be downloading all your software and sending you a paypal. Very nice work, Thanks a ton. Please keep working on it. Give us a speedfan type app but with your accurate measurements, logging and gui enhancements are all thats lacking.
Thanks again and my paypal is on the way
Thanks
Lee

[Kalway]

I just loaded it up on my 3200+ NC and it looks like core center is only a couple of degrees off from core temp under load. When it's idling core center is way off. Core temp reads ~48C under load while core center reads 50C. While idling Core temp reads like 40C while core center reads like 36C.

Want a GCPUID dump for it?

This is with watercooling, but it's like 90F in my room right now so it's not like I'm going to get awesomesauce temps right now.

A nice feature, if possible, would be an adjustable high temp warning.

[dinos22]

I did try the program with my mach II this post was on page 5 but since i didnt have any screen shots i guess no one read it.
my temps are listed in the text below
Lee

i like to look at pretty pics

yeah it seems there's a big problem with subzero temp readings

[lross78550]

here is a screen shot sorry for the size i run at 1920 x 1200 and anything smaller and the txt wasnt readable.
Lee

[NickS]

Looks very accurate

[dinos22]

yeah for idle that's fair enough............do load again

[DPGX]

Nice Util. My system for some reason isnt displaying the TDP or TCaseMax. Ive got an MSI K9N Platinum and an Orleans 3000+ Single core Athlon 64. Keep up the good work!

[One_Hertz]

kinda works...




Except for core 2 temps ALWAYS ghosting my evap temps, less than 3 degrees off at all times. The system was on single prime load. On idle core 1 gets -34 and core 2 something like 198C.

[high5]

Idle:
Attachment 49061

Load:
Attachment 49062

GCPUID dump attached.

[billdavis]

could you review this dump.txt to see if everything appears to be ok
________
PENNY STOCKS TO BUY

[soundx98]

Well CoreTemp and MBM5 are definately reading too different things.
CoreTemp appears to be reading CPU Diode and MBM5 (DFI version) is reading CPU Temp.
Using different programs to "load" the CPU also seems to make a difference between the two readings.
I'm not sure what program Pyr0 was using but I AM getting a large difference dpending on which program I use. (Thanks Pyr0)
Using SP2004 the difference was minimal (couple of degrees as I posted earlier).
Using SuperPi mod 1.5 the difference was much greater
Same thing when I used the System Stability Test in Everest Ultimate

I did some screenshots using Everest Ultimate v3.00.645 BETA.
(I would think the TRIAL will give you the same info)
I enabled On Screen Display and monitored CPU Temp AND CPU Diode in it.
Then set Everest to "poll" every 5 secs (same as MBM5 DFI config file). Took a bit to somewhat sync the two.

MBM5 - Core Temp - Everest



(I'm still amazed at how useful Alex Van Kaam's MBM5 is since development was stopped over 2 years ago. I've tried to color code the readings)

I than ran OCCT and Core Temp read a constant MINUS 49 degrees (wonder is it can read any lower? - I assume it is not working when OCCT is running))
As always MBM5 dashboard is not visible when running OCCT, but Everest Ultimate displayed just fine.

OCCT - CoreTemp - Everest Ultimate



Everest Ultimate has a Program under "Tools" called System Stability Test.
It stresses the CPU to 100% and you can see the Vista Sidebar reflecting it at 100% also.

Everest System Stablity Test - CoreTemp - MBM5 - Everest OSD



I'm gonna try testing with different OSes again and different methods of "loading" the CPU. Willing to bet that Vitaliy Jungle's "Next Sensor" Program is going to give some pretty similar results. Hopefully I can get my hands on a dual core rig over the next two days to do more tests with.

[Nedjo]

woah! How is this posible!

http://www.benchmark.co.yu/forum/sho...&postcount=229

Guy says it's Prime stable!

[MikalCarbine]

Where'd you get the .645 Beta of Everest?? I can only find the .630

[The Coolest]

Well CoreTemp and MBM5 are definately reading too different things.
CoreTemp appears to be reading CPU Diode and MBM5 (DFI version) is reading CPU Temp.
Using different programs to "load" the CPU also seems to make a difference between the two readings.
I'm not sure what program Pyr0 was using but I AM getting a large difference dpending on which program I use. (Thanks Pyr0)
Using SP2004 the difference was minimal (couple of degrees as I posted earlier).
Using SuperPi mod 1.5 the difference was much greater
Same thing when I used the System Stability Test in Everest Ultimate

I did some screenshots using Everest Ultimate v3.00.645 BETA.
(I would think the TRIAL will give you the same info)
I enabled On Screen Display and monitored CPU Temp AND CPU Diode in it.
Then set Everest to "poll" every 5 secs (same as MBM5 DFI config file). Took a bit to somewhat sync the two.

MBM5 - Core Temp - Everest



(I'm still amazed at how useful Alex Van Kaam's MBM5 is since development was stopped over 2 years ago. I've tried to color code the readings)

I than ran OCCT and Core Temp read a constant MINUS 49 degrees (wonder is it can read any lower? - I assume it is not working when OCCT is running))
As always MBM5 dashboard is not visible when running OCCT, but Everest Ultimate displayed just fine.

OCCT - CoreTemp - Everest Ultimate



Everest Ultimate has a Program under "Tools" called System Stability Test.
It stresses the CPU to 100% and you can see the Vista Sidebar reflecting it at 100% also.

Everest System Stablity Test - CoreTemp - MBM5 - Everest OSD



I'm gonna try testing with different OSes again and different methods of "loading" the CPU. Willing to bet that Vitaliy Jungle's "Next Sensor" Program is going to give some pretty similar results. Hopefully I can get my hands on a dual core rig over the next two days to do more tests with.

The thing is i see OCCT uses the same type of driver that Core Temp uses. But core temp dynamically loads and frees the driver (enables a user to run multiple instances of the program). OCCT overtakes it and when Core Temp tries to read the temperature and fails, reports a '0' to the function that calculates temp and as the temp calculated by "Temp == Value - 49" this results in 0-49 = -49C displayed.
I think I'm going to rename the driver when I get the chance, this way should not conflict with any other software.

could you review this dump.txt to see if everything appears to be ok

Looks just fine.

woah! How is this posible!
http://www.benchmark.co.yu/forum/sho...&postcount=229
Guy says it's Prime stable!

First of all there's a thread in the Intel section as well, I suggest you post this there.
Second, the B0 rev seem to be reporting some crazy temps via the DTS. The few B1s that I saw, on the other hand report temps that make sense and seem to be at least somewhat accurate.

[soundx98]

found a link to 3.00.645 in a post here
http://www.lavalys.com/forum/index.php?showtopic=1517#
Chipset temp added to tool tray.

Thank you for the information TheCoolest, very nice work