"Core Temp" - CPU Temperature monitoring tool for the K8 series

The Coolest · Jul 8, 2006

VID is not VCore, VID is the setting for voltage that is selected in the CPU (can be controlled via CoolnQuiet feature, even if its disabled in the BIOS). VCore is the actual voltage that mobo sensors read and report to other software. Core Temp reads the VID setting from the CPU, and not the actual voltage.

Regarding TDP click the "AMD64 TCaseMax - Find your A64/Opteron's Max Power Draw!" link in my sig, the AMD64 TCaseMax program was integrated into Core Temp, and is now discontinued, but all the explanation is there.

davidk21770 · Jul 8, 2006

Followup to the 75/55C under prime95.
I let it run until one of the cores failed (didn't used to at this clock, etc

, but that's why I'm looking at programs like this). But, it was the COOL core that failed (the one at 55, not the one at 75 -- and they went to 80/60 after a bit of priming)! I'm thinking that it's inaccurate, at least for my CPU.

I wonder if this is why the MB manus aren't using it? It's not really ready for prmie time yet?

The Coolest · Jul 10, 2006

Update:

Version 0.9.0.91 - 5th July, 2006
- Logging feature added - Just copy into Excel for easy graphing.
- Fixed a bug where some programs would cause the program's driver to fail.

*EDIT:* There was a problem with this latest version crashing. It should be fixed now. So please redownload

chaos · Jul 11, 2006

My core#1 is about 10c cooler than my core #2..And Everest is reading at the higher temp... Is that normal and can I do anything about that???

davidk21770 · Jul 12, 2006

Same problem here -- 12-15C difference. And, this is with the same chip in 2 different MBs (same board, different rev) and with the waterblock removed and reapplied many times thinking that it might be a poor joint or that I didn't get the AS5 centered. I'm considering removing the IHS, it's just a tough decision with such and expensive chip -- I'm working up my courage

. At least the DFI Expert uses a low profile CPU socket so I don't have to dremmel anything!

Interestingly enough, it's the COLD chip that always fails prime95 1st making me wonder if it's really a temperature offset or a calibration issue. Only way to know is to get the top off <sigh>.

Hipcrostino · Jul 12, 2006

Hi, Just wanted to point somthing out.

If you look at both my and Niku-Sama attached pictures you will note that our Venus opty 14x cores are detected at San Diego cores while cpuz detects it correctly. Also look at my mbm5 temps and my core #1 temps in your program. it would be sweet if yours was correct. my colling is Air and i would say ambient is around 17-16 degrees.

p.s I just want to point out that although some of you with dual core procs may be running only one core under load the other will heat up too simply due to the close proximity to the working silicone it has.

p.p.s I think it is the correct core temp reported, as its jumping up and down constantly (as it should). its moving between 14 and 18*c

p.p.s.s WOW! just ran a super pi 1M. temp went from 18 to 27 degrees in one hit! maintained 26 - 29 under load. (for 33 seconds)

Scar · Jul 12, 2006

davidk21770 said:
Same problem here -- 12-15C difference. And, this is with the same chip in 2 different MBs (same board, different rev) and with the waterblock removed and reapplied many times thinking that it might be a poor joint or that I didn't get the AS5 centered. I'm considering removing the IHS, it's just a tough decision with such and expensive chip -- I'm working up my courage .

Don't bother, mine does the same thing and the ihs is removed.

davidk21770 · Jul 12, 2006

Scar said:
Don't bother, mine does the same thing and the ihs is removed.

Thanks. I'll relax now

.

The Coolest · Jul 12, 2006

The readings aren't perfect. the DTS are most likely not calibrated that well as this feature was pretty much undocumented on Pre-RevF CPUs

Hipcrostino · Jul 13, 2006

question is which temp do i trust. It really dosn't matter atm, but come summer the ambient temps a going to rise and i might have temp issues

CanadianTSi · Jul 13, 2006

If your cores have a big difference in temps (5-10 degree's or more) then theres a good chance your not making good contact with one of them and you should try re seating your heatsink...it's worth a shot.

Here's my Naked X2 at Idle and Full Load.

QuietIce · Jul 13, 2006

CanadianTSi said:
If your cores have a big difference in temps (5-10 degree's or more) then theres a good chance your not making good contact with one of them and you should try re seating your heatsink...it's worth a shot.

That might work if the IHS is off but a lot of us are just not ready to take the plunge, yet ...

CanadianTSi · Jul 13, 2006

QuietIce said:
That might work if the IHS is off but a lot of us are just not ready to take the plunge, yet ...

I did it for the first time the other day, its really not that hard. Took me maybe 5 minutes.

deathman20 · Jul 14, 2006

Very nice so thats how the temp gets calculated for the CPU. Take the 2 temps and divide it.

Intresting.

QuietIce · Jul 14, 2006

deathman20 said:
Very nice so thats how the temp gets calculated for the CPU. Take the 2 temps and divide it.

Interesting.

It's apparently not that simple. Running Prime95 on core 1 my readings are Core 0: 53, Core 1: 29 but the CPU temp reads 35 - go figure ...

deathman20 · Jul 14, 2006

QuietIce said:
It's apparently not that simple. Running Prime95 on core 1 my readings are Core 0: 53, Core 1: 29 but the CPU temp reads 35 - go figure ...

Oh well.

BTW Sweet works on the my Core Duo work laptop

38C Idle ish. 100C Tjunction wow.

The Wicker Man · Jul 15, 2006

Ok, I read threw this thread, and still am not sure who to trust, Speed Fan or Core Temp, I want to trust core temp......

TeaBoy · Jul 17, 2006

I'll throw my oar in the water and test this utility. BTW, nice work!

Both MBM5 and Speedfan are reading my rigs core temps the same so here's the comparison with the Core Temp beta utility under different load situations, (using 2 x Prime95).

Core 0 Idle / Core 1 Idle = (Core 1 36C) (Core 2 35C) (MBM5/SF 39C)
Core 0 Load / Core 1 Idle = (Core 1 40C) (Core 2 42C) (MBM5/SF 45C)
Core 0 Idle / Core 1 Load = (Core 1 48C) (Core 2 44C) (MBM5/SF 45C)
Core 0 Load / Core 1 Load = (Core 1 50C) (Core 2 48C) (MBM5/SF 48C)

Slightly lopsided but that may very well be the way I applied TIM or the way the heatspreader was factory fitted. Also it looks like the cores temps are reversed unless the naming is incorrect.

BTW, I think you should name the cores 0 & 1 rather than 1 & 2 as most programs use the 0 & 1 naming and it might be a bit less confusing.

Not sure if that helps.

deathman20 · Jul 18, 2006

Is there a way to turn off the tempature logging? It creates a file every few seconds and definatly uses power doing it.

Silverfoot · Jul 18, 2006

Awsome proggy. Seems that something is whacked somewhere though..

I'll post the exact diff. tonight, but out of memory there is a BIG gap between this and MBM

2x F@H __ MBM= 31 //CT=48

(roughly~yikes)

"Core Temp" - CPU Temperature monitoring tool for the K8 series

Senior Member

Member

Senior Member

Member

Member

Member

Attachments

Member

Member

Senior Member

Member

Member

Disabled

Member

High Speed Premium Senior

Disabled

High Speed Premium Senior

Member

Member

High Speed Premium Senior

Member

Similar threads