"Core Temp" - CPU Temerature monitoring tool

[Ace-a-Rue]

I'm having the same issue on an Abit 35 pro with 2 different 9650's. I am going to try them in a new board tomorrow and see if anything changes. BTW, my Q6600's temps are within 1-2 of each other on this board.

what's your pack date?...also, do you have core #0 locked in at 51C at idle?

which bios are you using on the Pro board with your 9650?...i hope it is the 16.b04 beta bios...that worked for me when i had the board up and running.

[St3vil_68]

1 is L742A753 pack 12/07
Other is L739A770 Pack 12/07

[St3vil_68]

16.b04 beta bios and no temps are "locked".

[The Coolest]

So a BIOS update "fixed" this?

[St3vil_68]

So a BIOS update "fixed" this?

No, at least not the 16.b04.

[Ace-a-Rue]

So a BIOS update "fixed" this?

no, it did not...i have the IP35 Pro in the closet, at the moment...when i was using the Pro board with the 9650 and the beta 16.b04, my 9650 still had a locked 51C on core #0.

[JohnZS]

One of my ideas was that maybe each core had it's own Tjunction, but looking at the dump tells me that they all have a Tjunction Max of 105C.
I don't know why there is such a difference between cores.
Either a poor callibration of the DTS or maybe bad contact between the cores and the IHS (just shooting in the dark)

Thanks for looking Arthur
What does it mean if there is a bad calibration of the DTS?
would this be fixable with a BIOS update or a Core Temp Update?
Also another thing, why is there such a huge difference between core temp and ASUS probe?
ie ASUS Probe reports CPU @ 16C yet coretemp reports 31 31 52 52.
Is this because ASUS have a BIOS problem?
John

[Ace-a-Rue]

i'd think DTS is a core problem versus a bios problem.

[The Coolest]

I think so too.
Could be defective\poorly calibrated DTS.

Calibration is done at the Intel fab and BIOS updates can't fix this as its individual for every CPU.

[JohnZS]

I think so too.
Could be defective\poorly calibrated DTS.

Calibration is done at the Intel fab and BIOS updates can't fix this as its individual for every CPU.

Oh no

That is disapointing, so basically I am trapped with a chip which either has 2 cores running exceedingly hot or 2 defective TDiodes?

IMHO that shouldn't happen on a £600 CPU!

Any suggestions?

Shall I try for an RMA?

John

[Ace-a-Rue]

Oh no

That is disapointing, so basically I am trapped with a chip which either has 2 cores running exceedingly hot or 2 defective TDiodes?

IMHO that shouldn't happen on a £600 CPU!

Any suggestions?

Shall I try for an RMA?

John

are you beyond the return policy with your seller for defective hardware?...if so, then the only way to get an exchange is to call Intel tech support...i tried but i got the stone-wall treament from a guy who is in costa rica...he spoke very good english but he wasn't going to give an inch in allowing me an RMA.

i was able to return my 9650 to the online retailer since i was one day before the 30 day policy expired...shew!...should have a new one in 5-7 days....i am hoping it is not the same pack date.

[fgw]

One of my ideas was that maybe each core had it's own Tjunction, but looking at the dump tells me that they all have a Tjunction Max of 105C.

hi art,

long time no hear - i have not been around for some time.

was reading the last few pages of this thread and your statement made me to jump on this train.

i'm very interested in the functionality around DTS and coretemp measurements. in the old days of e6400 about a year ago the DTS reported temps as an offset to TCC activation threshold.

has this changed for newer processors? is DTS now reporting the offset to Tjmax and not TCC activation anymore?

also, in the past it was impossible to read TCC activation threshold through any registers. reading your statement i get the feeling Tjmax can be read in an MSR - is this correct?

could you please give me a few words on this? if you prefer, you can also pm me. as i'm registered over on the ALCPU Forums you can find my email there if you don't have it anymore anyway.

cheers

[JohnZS]

are you beyond the return policy with your seller for defective hardware?...if so, then the only way to get an exchange is to call Intel tech support...i tried but i got the stone-wall treament from a guy who is in costa rica...he spoke very good english but he wasn't going to give an inch in allowing me an RMA.

i was able to return my 9650 to the online retailer since i was one day before the 30 day policy expired...shew!...should have a new one in 5-7 days....i am hoping it is not the same pack date.

Thanks I've got an RMA number and they are collecting it on Wednesday, I am packaging it all up afterwork tomorrow. I'm not sure when I will get my replacement, but they had no problems with the reason I gave as it could be a potential for serious damage.

I'm not sure if it is related but during video encoding or wprime after 30 mins Core3 starts to go mad and I "lose" it's sensor (it goes blank) and the wprime thread laggs for core 3 by 3%! (after 30mins) and video encoding slows down. It's as if I "lose" a core.

Hopefully I will get a replacement within 2 weeks which as less then 5C spread between all 4 cores.

John

[Ace-a-Rue]

glad to hear that you got an easy resolution.

supershanks in the abit IX38 thread told me he got a bad E6600 or E6400...he thought the board was bad and therefore, he replaced that first...same problem...not until he took the cpu back and got a replacement did everything run smoothly.

[JohnZS]

glad to hear that you got an easy resolution.

supershanks in the abit IX38 thread told me he got a bad E6600 or E6400...he thought the board was bad and therefore, he replaced that first...same problem...not until he took the cpu back and got a replacement did everything run smoothly.

Thanks Ace-a-Rue
It's all packaged up now, so I'm back on my old S478 P4 now.
I should get my replacement nextweek. (providing they are still in stock).
I included a brief resumé of the problem and a picture of coretemp and taskmanager just in case the "testers" just pop in the CPU and say "Oh it POSTS therefore nothing is wrong".

I'll have a dig through that thread to see how supershanks turned out before and after RMA.
Thanks
John

[Ace-a-Rue]

i'm getting this sick feeling that intel is having quality control issues.

[JohnZS]

i getting this sick feeling that intel is having quality control issues.

I hope not, but I sort of feel the same as our pack dates were the same (11/21) Perhaps the damage MIGHT have been caused in transit from the Intel Factory?

Or because 45nm process is still in it's early days...perhaps a few teething problems as they ramp up the process?

John

[Ace-a-Rue]

the 2nd sentence is more like it.

[PaRn ssR]

Thank

[Ace-a-Rue]

replacement 9650 arrived today with a pack date of 12/07/2007...i'm hoping that the core temps are within +/- 3-4C of each other.

[JohnZS]

replacement 9650 arrived today with a pack date of 12/07/2007...i'm hoping that the core temps are within +/- 3-4C of each other.

That's good news, I hope your temps and everything work out as they should do
I will not be getting my replacement until next week, the fustrating thing is that the courier collected it on Wednesday but they have still not received it yet as it is at a sorting office. Hopefully they will receive it just before the weekend.

John

[Ace-a-Rue]

here is the new 9650 at stock speed...temps are more reasonable....posted in another thread so the link is provided:

http://www.xtremesystems.org/forums/...3&postcount=32

[St3vil_68]

here is the new 9650 at stock speed...temps are more reasonable....posted in another thread so the link is provided:

http://www.xtremesystems.org/forums/...3&postcount=32

That's great Ace, but I have 2 with same pack date, 1 reads temps about like your new one and 1 like your old. I'm beginning to think there may be some QC issues at Intel...

[The Coolest]

Looks good

[gmod]

Just got my 9650 going tonight...Using fusion 120.3 to cool just the cpu...

Here are my temps
32/27/27/29 1.49V Windows...

Cpu is a 39A 12/07