"Core Temp" - Official Developer Thread.

The Coolest

Senior member
Dec 16, 2004
364
54
91
www.alcpu.com
Download Core Temp v1.18.1

This program lets you monitor Intel's Core, Xeon, Pentium, Celeron and Atom series, AMD Ryzen, APU, FX, Phenom, Athlon64 and later VIA C7 and Nano chips' die temperature.
The temperature readings are very accurate as the data is collected from a Digital Thermal Sensor (or DTS) which is located in each individual processing core, near the hottest part.
For more information click here

There is also an article posted about Core Temp and a little background on standard thermal sensors - Article
Please help out Core Temp popularity by Digging it.

If you find any bugs, please post it in the "Bugs" section ASAP!
Any suggestions on improvements or ideas are welcome in the "Development" section.
Note 1: Please read the ReadMe!.txt file for some useful tips.

Are you a developer yourself and have ideas for new features?
Click here for information about developing your own plugins for Core Temp.

What's new:

Version 1.18.1 - 24th September, 2023


- Fix: Do not display TDP and TjMax for Zen and Zen+ CPUs
- Fix: Incorrect BCLK on legacy (Pre-Zen) AMD CPUs
- Fix: Check for updates feature

Version 1.18 - 4th December, 2022

*** Core Temp Monitor is now available, Windows Phone and Android apps apps to monitor your machines from anywhere in the world!
More information is available here


- New: AMD Zen 4, Zen 3 APU and Zen 2 APU support
- New: Intel Raptor Lake and Alder Lake support
- New: TDP, TjMax, multiplier range detection on desktop AMD Zen platforms

- Fix: Address the kernel-mode driver vulnerability/exploitation issues
- Fix: Redesign Bclk detection on all AMD platforms starting from the Phenom series
- Fix: Workaround the Bclk detection issues on Intel Skylake and newer series
- Fix: Bclk detection on older Intel platforms, utilizing x2Apic configuration
- Fix: Thread count on Intel hybrid architectures (Raptor/Alder Lake)
- Fix: Improve support for multiple older AMD and Intel processors
- Fix: Base multiplier detection on AMD Zen processors
- Fix: Incorrect temperature on AMD Zen processors (missing -49C offset)
- Fix: Engineering Sample recognition on newer Intel processors
- Fix: Multiple minor bugs

- Change: Windows Vista 64-bit is no longer supported


Older versions
 
Last edited by a moderator:

The-Noid

Diamond Member
Nov 16, 2005
3,117
0
76
Really a good program The Coolest. I just have one question about Core 2 Duo that maybe you know the answer to. On phase or any Sub-Zero cooling solution the temperature reported by Core Temp or Everest is different for each CPU. It seems to reach a certain low and stays there. With one of my C2D's it was 2, the next 8, another 13, and finally 4. They were all the same model E6700, and regardless of voltage used or what the cpu temperature was it just seemed to bottom there. Do you think this dictates that the DTS is poorly calibrated or just not made for Sub-Zero operation?

The only question is begs to ask is if the processor bottoms at 13 does that mean when it is in use you can subtract 13 degrees from the core temp, or 2 or 4. Makes me wonder if there is something I am missing.
 

The Coolest

Senior member
Dec 16, 2004
364
54
91
www.alcpu.com
I'm not sure why this happens, but I know for a fact this happens on all Core and Core 2 CPUs.
I have no clue as to why it happens.
I'm guessing that the DTS was not really designed to work properly under extreme conditions.
 

The-Noid

Diamond Member
Nov 16, 2005
3,117
0
76
Do you think in that case that 13 is calibrated to 0, or am I fishing for answers as to why Core Temps were 20 degrees higher then what P5WDH was reading on that processor. I believe that the processor is definetely reading the DTS correctly, I just wish Intel possibly could have done a better job calibrating these before they went out.
 

The Coolest

Senior member
Dec 16, 2004
364
54
91
www.alcpu.com
Not sure what you mean by that.
I'm sure that Intel has callibrated these DTS to be as accurate as possible.
Also the DTS is placed in the hottest part of the core, that is probably what really makes Core Temp report a much higher temp on these chips than mobo sensors.
 

The-Noid

Diamond Member
Nov 16, 2005
3,117
0
76
I was thinking its possible that the DTS reads a temperature 0-255 or 0-254. So at a certain point you hit a negative number that is below 0 and it comes back with 13 or 2 or 4. Is it possible that on that chip 13 is 0. So if you subtract 13 from temperature readings you have your real temperature. Just something I was kind of thinking of. Obviously the DTS can not read a negative number. So it reaches the bottom of its temperature range which to me would be 0, and reports back 13. Maybe the DTS is very accurate but the data it is reporting is not.

So if the farthest you could go is 13 and the temperature reading was 58, 58-13 = real temperature.

It is just that this chip always seemed really high in setup compared to the others in Core Temp. Motherboard showed it in a normal range, then when I put it on phase it was 13 where the others on Phase were single digits. I started to think that possibly had something to do with the higher temperatures.
 

OcHungry

Banned
Jun 14, 2006
197
0
0
Hi coolest.
Can you make one for Win Vista? The 32bit Vist will work if I run the program compatibilty after a few trials, but havnt been able to make it work in 64bit Vista.
Is there a way to right the codes in 64bit or makes it 64bit ready?
thanks
 

The Coolest

Senior member
Dec 16, 2004
364
54
91
www.alcpu.com
Yoxxy:
I really doubt that is what happening. Either way the DTS can theoretically return a value between 0 and 127, and it just gets "stuck" when temps go low enough.
I'm pretty sure its just a DTS limitation and not a miscallibration of some sort.
Again, you see higher temps because the DTS reports more accurate, but also temperatures that are taken from the HOTTEST parts of the core.

OcHungry:
Unfortunatly as smart as Microsoft is now there is no way to make Core Temp work on Vista x64, as the driver needs digital signing. And the license costs $500/yr last time I checked.
I'm pretty sure it crashes on x64 Vista, right? I don't have a PC with good enough specs to install Vista on and debug it (to at least get the GUI up and running).
 

lopri

Elite Member
Jul 27, 2002
13,209
594
126
Without this little prog I wouldn't have achieved the OC that I currently have! Thank you for your work. One question: Is there any other info you can get from the DTS than each core's temperature? Just wondering.

That license thingy for an app like this is just ridiculous. With the rumoured one-time transfer rule in the EULA, I doubt I will get the Vista anytime soon. I can wait till MS gives up the stupid idea and price drops. By that time SP2 should be out, too. :D
 

lopri

Elite Member
Jul 27, 2002
13,209
594
126
Also would like to know how accurate the readings are for A64 CPUs. (Where is the sensor?)
 

The Coolest

Senior member
Dec 16, 2004
364
54
91
www.alcpu.com
Cuurently that's all it reports. There is also an overheat flag that Core Temp can detect and alarm you.

*EDIT:*
I'm not sure about where the DTS is placed exactly, I'm pretty sure its the THE HOTTEST part, but it isn't the coolest part either.
I don't know how accurate the DTS is on pre-RevF chips as this feature was unofficial, but on RevF chips its supposed to be quite accurate as well
 

lopri

Elite Member
Jul 27, 2002
13,209
594
126
Bump for the excellent article and, once again, the excellent little toy.
 

gpgofast

Senior member
Oct 6, 2000
351
0
0
On my new FX-55 w/ an aBit AN832X board at stock voltage and clocked at 2.8GHz, the uGuru utility reads about the same idle temp-32 C, but fully loaded, uGuru reads 50 C while Core Temp reads 43 C. If I understand corectly, the Core Temp reading is more accurate???
 

homestarmy

Diamond Member
Apr 16, 2004
3,528
0
0
artwilbur.com
This seems like a pretty neat program... but one issue.

On my desktop, with an Athlon X2 3800+ @ 2.4GHz, it says Core 1 is CONSTANTLY much hotter than Core 0.

Even when running Prime95 torture test on Core 0, Core 1 is still hotter by 5C. At idle, it is around 6C hotter. Any idea why this is?

On my Core Duo laptop, they are the same, give or take.
 

The Coolest

Senior member
Dec 16, 2004
364
54
91
www.alcpu.com
gpgofast:
It is supposed to be.
Its up in the air whether or not it actually is.
In most cases I've seen Core Temp has been pretty accurate. especially if its a Rev E chip, but many CG chips gave believable results as well.

homestarmy:
This is a known fenomena. I only saw one person who fixed this by removing the IHS, but many users with removed IHS still reported no change.
I have no idea why that's happening, it could be a sign of poor calibration, or just a classic "no chip is create equal", maybe?
 

homestarmy

Diamond Member
Apr 16, 2004
3,528
0
0
artwilbur.com
BTW, when do you plan on making an icon to sit in the system tray? IMO anything is better than the generic one... it makes me feel like something is broken when I see it in my tray...

Thx :).
 

The Coolest

Senior member
Dec 16, 2004
364
54
91
www.alcpu.com
Yeah I know what you mean.
I can't draw, especially small stuff like icons. But one user already submitted me an icon. I might use it, its not perfect but looks much better than the stock one.
I also hope for the next release to have an actual temp shown in the system tray.
 

Munky

Diamond Member
Feb 5, 2005
9,372
0
76
This app doesnt really measure the cpu voltage, does it? Because it says 1.35v for my Opty 165, even though I have it set at 1.43v.
 

homestarmy

Diamond Member
Apr 16, 2004
3,528
0
0
artwilbur.com
Originally posted by: The Coolest
Yeah I know what you mean.
I can't draw, especially small stuff like icons. But one user already submitted me an icon. I might use it, its not perfect but looks much better than the stock one.
I also hope for the next release to have an actual temp shown in the system tray.

That's nice, thanks!

And I read in the Dig link why my X2 temps are off... they're basically "unofficial" and not guaranteed to be accurate, or maybe guaranteed to be not acurate :).