![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
-
Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.
Which core is which?
- Thread starter todlerix
- Start date
Good timing. I finally noticed a problem with what most temperature software is reporting for Quad core processors. There is a problem with the data for some cores getting swapped so what software reports doesn't reflect the physical core order.
I think the latest beta version of RealTemp is correctly handling this problem. You can download version 2.69.1 here:
http://www.fileden.com/files/2008/3/3/1794507/RealTempBeta.zip
You need to unzip this file and copy the new RealTemp.exe into your present RealTemp folder and overwrite the old version.
In the Settings window it will show you your core order. Most software assumes 0 1 2 3 but by the looks of your temps I think RealTemp is going to show that a couple of cores, maybe the center two cores, have been swapped. RealTemp automatically tries to report your temps in the correct physical order now. This core swapping can happen any time after a reboot. My Q6600 was the same for 3 months until one day the center cores swapped places and it has been swapped for about a month now.
If you don't have any sensors that are stuck then you could read the docs and do the calibration procedure. That will help balance your idle temps. At full load when running Prime, a difference between the two sets of cores of 5C is typical. With my CPU fan turned off I've seen a difference as high as 9C between the two sets of cores but it is more like 3C when my Tuniq fan is on high while running Prime.
If you need some more info then open up the About... box in RealTemp and send me an e-mail. I know it can be a challenge getting accurate temps out of a 45nm Quad.
Edit: The new RT also has a Mini-Mode feature. Just double click on the GUI to enter or exit this mode.
I think the latest beta version of RealTemp is correctly handling this problem. You can download version 2.69.1 here:
http://www.fileden.com/files/2008/3/3/1794507/RealTempBeta.zip
You need to unzip this file and copy the new RealTemp.exe into your present RealTemp folder and overwrite the old version.
In the Settings window it will show you your core order. Most software assumes 0 1 2 3 but by the looks of your temps I think RealTemp is going to show that a couple of cores, maybe the center two cores, have been swapped. RealTemp automatically tries to report your temps in the correct physical order now. This core swapping can happen any time after a reboot. My Q6600 was the same for 3 months until one day the center cores swapped places and it has been swapped for about a month now.
If you don't have any sensors that are stuck then you could read the docs and do the calibration procedure. That will help balance your idle temps. At full load when running Prime, a difference between the two sets of cores of 5C is typical. With my CPU fan turned off I've seen a difference as high as 9C between the two sets of cores but it is more like 3C when my Tuniq fan is on high while running Prime.
If you need some more info then open up the About... box in RealTemp and send me an e-mail. I know it can be a challenge getting accurate temps out of a 45nm Quad.
Edit: The new RT also has a Mini-Mode feature. Just double click on the GUI to enter or exit this mode.
Monkey God
Mangina Full of Sand
- Joined
- May 7, 2007
- Messages
- 6,723
this qx6700 pic shows they are left to right, and the "bottom" is cache
caveat: I am not a CPU expert
http://www.neoseeker.com/Articles/Hardware/Reviews/core2quad_qx6700/
caveat: I am not a CPU expert
http://www.neoseeker.com/Articles/Hardware/Reviews/core2quad_qx6700/
Core 0 / core 1 are on the left and core 2 / core 3 are on the right.
Why not run Prime small FFTs on all 4 cores and the latest version of RealTemp. I think RT will organize your cores so the two hot ones are 0/1 and the two cools ones are 2/3. I'm not really sure why but I think most Quads are similar. The 45nm dual cores have a very small real difference between each core but there are so many 45nm Quads with sensor issues at the temperatures that most people run them at that it can be hard separating sensor / reporting issues from actual issues with different temperatures.
One theory I have is that the two cores on the right, core2 / core3, typically make better contact with the IHS so heat transfer is better and they tend to run cooler. If I can find an official Intel document that shows the core order I'll post a link to it.
Thanks Monkey God. The picture on page 12 of that article is the one that I was looking for:
http://www.neoseeker.com/Articles/Hardware/Reviews/core2quad_qx6700/12.html
Why not run Prime small FFTs on all 4 cores and the latest version of RealTemp. I think RT will organize your cores so the two hot ones are 0/1 and the two cools ones are 2/3. I'm not really sure why but I think most Quads are similar. The 45nm dual cores have a very small real difference between each core but there are so many 45nm Quads with sensor issues at the temperatures that most people run them at that it can be hard separating sensor / reporting issues from actual issues with different temperatures.
One theory I have is that the two cores on the right, core2 / core3, typically make better contact with the IHS so heat transfer is better and they tend to run cooler. If I can find an official Intel document that shows the core order I'll post a link to it.
Thanks Monkey God. The picture on page 12 of that article is the one that I was looking for:
http://www.neoseeker.com/Articles/Hardware/Reviews/core2quad_qx6700/12.html
chrisf6969
[H]F Junkie
- Joined
- Oct 27, 2003
- Messages
- 9,011
could just be bad readings on those 2 cores.
Try lapping your IHS and heatsink.
Try lapping your IHS and heatsink.
- Joined
- Feb 3, 2004
- Messages
- 26,937
The one reporting the lowest temperature is most likely a stuck sensor , Its a common problem on the Q9450's
hafaznmatt
Limp Gawd
- Joined
- Jul 7, 2007
- Messages
- 352
It could just be faulty readings coming from the sensors. Try different system specs, and different loads.
The accuracy of RealTemp can be improved with some basic adjustments but it looks like your CPU is a little beyond that. There is still hope but I will have to create a custom calibration for your CPU.
Core0 / core1 are moving similar amounts from idle to full load and so are core2 / core3. This is a good sign that the sensors are not significantly sticking. As long as they're not sticking, it's possible to get some reasonably accurate temperatures out of them. The final result won't be perfect but it will probably be much more accurate and better balanced than the random numbers you have now.
If you would like me to give this a try then you need to send me some data. Turn on the RealTemp logging feature and set the interval to 1 second so I will have plenty of numbers to look at.
At your normal speed of 3200 MHz, run your processor idle for about 1 minute and then run Prime95 small FFTs on all cores for about 3 minutes. Stop that and return your computer to idle for the final minute. You don't have to be exact with the amount of time. I won't complain if you send me too much data! When done you can go into your RealTemp folder and attach the RealTempLog.txt file to an e-mail.
Next go into the bios and set your CPU to default. 333 MHz and lock it at a 6.0X multi. Lower the core voltage down to approximately 1.10 volts. If you think setting your computer like this might screw something up then you can also set the MHz, core voltage and multiplier to AUTO and enable C1E in the bios. As long as your motherboard is working properly, this should accomplish the same thing where your multiplier and core voltage will both automatically drop when idle in Windows.
Boot up and let your computer idle for about 5 minutes or until the temps have stabilized. Send me a screen shot showing CPU-Z / RealTemp of this. Remember to open up the Settings window in RealTemp so I can see that the core order is 0 1 2 3. During this part of the test, the goal is to get your CPU as cool as possible so crank up the fans and open the case if you can. Also include your room temperature.
E-mail this info to me:
Real_Temp@yahoo.ca
It might take me a few days to have a look at things and come up with a solution for you to try. Just keep sending e-mails to me in case I forget. A [H]ard Gawd with 5.2 years experience deserves some decent temps out of his new Quad!
Core0 / core1 are moving similar amounts from idle to full load and so are core2 / core3. This is a good sign that the sensors are not significantly sticking. As long as they're not sticking, it's possible to get some reasonably accurate temperatures out of them. The final result won't be perfect but it will probably be much more accurate and better balanced than the random numbers you have now.
If you would like me to give this a try then you need to send me some data. Turn on the RealTemp logging feature and set the interval to 1 second so I will have plenty of numbers to look at.
At your normal speed of 3200 MHz, run your processor idle for about 1 minute and then run Prime95 small FFTs on all cores for about 3 minutes. Stop that and return your computer to idle for the final minute. You don't have to be exact with the amount of time. I won't complain if you send me too much data! When done you can go into your RealTemp folder and attach the RealTempLog.txt file to an e-mail.
Next go into the bios and set your CPU to default. 333 MHz and lock it at a 6.0X multi. Lower the core voltage down to approximately 1.10 volts. If you think setting your computer like this might screw something up then you can also set the MHz, core voltage and multiplier to AUTO and enable C1E in the bios. As long as your motherboard is working properly, this should accomplish the same thing where your multiplier and core voltage will both automatically drop when idle in Windows.
Boot up and let your computer idle for about 5 minutes or until the temps have stabilized. Send me a screen shot showing CPU-Z / RealTemp of this. Remember to open up the Settings window in RealTemp so I can see that the core order is 0 1 2 3. During this part of the test, the goal is to get your CPU as cool as possible so crank up the fans and open the case if you can. Also include your room temperature.
E-mail this info to me:
Real_Temp@yahoo.ca
It might take me a few days to have a look at things and come up with a solution for you to try. Just keep sending e-mails to me in case I forget. A [H]ard Gawd with 5.2 years experience deserves some decent temps out of his new Quad!