Dual Core Temps Difference?
- Sparky
- Legit Extremist
- Posts: 727
- Joined: Fri Feb 16, 2007 4:47 pm
- Location: Huntington Beach, CA USA
Dual Core Temps Difference?
Is it normal to have different temps on each core on a dual core cpu? I have an Intel E6700 and core 1 runs 4 degrees Celsius (or greater) than core 0 when I’m running Intel Thermal Analysis Tool (TAT). Makes me wonder if I need to remove and reinstall my heat sink and fan. Perhaps I don't have the artic silver spread evenly???
Has anybody else had this occur?
Has anybody else had this occur?
MSI Meg X570 Unify | Ryzen 3900X | Noctuna NH-U12A | G.Skill RipJaws DDR4 3600 16GB | Samsung 970 Pro 512GB M.2 NVMe | (2) WD Blue 3TB Backups | MSI 2060 Super Gaming X | Meshify - C | SB X-Fi
-
- Legit User
- Posts: 22
- Joined: Sun Feb 25, 2007 5:17 am
- Sparky
- Legit Extremist
- Posts: 727
- Joined: Fri Feb 16, 2007 4:47 pm
- Location: Huntington Beach, CA USA
Thanks. I have noticed on task manager that CPU 1 seems to show more use than CPU 0, so perhaps this is the reason.
MSI Meg X570 Unify | Ryzen 3900X | Noctuna NH-U12A | G.Skill RipJaws DDR4 3600 16GB | Samsung 970 Pro 512GB M.2 NVMe | (2) WD Blue 3TB Backups | MSI 2060 Super Gaming X | Meshify - C | SB X-Fi
- kenc51
- Legit Extremist
- Posts: 5167
- Joined: Thu Jun 23, 2005 1:56 pm
- Location: Dublin, Republic of Ireland
- Contact:
He's using a E6700 not a Pentium D. The C2D chips are two cores on one die! (ie. Native dual core)Zoogle wrote:Also, Intel's dual cores are technically two dies on a single chip. So one die could have more voltage leak than the other one.
It's normal for each core to have different temps, even under full (equal) load. I wouldn't worry about it!
- Sparky
- Legit Extremist
- Posts: 727
- Joined: Fri Feb 16, 2007 4:47 pm
- Location: Huntington Beach, CA USA
Thanks again for all the feedback. I also have an AMD 4800X2, but I've never able to verify the temps on individual cores utnil I used TAT.
MSI Meg X570 Unify | Ryzen 3900X | Noctuna NH-U12A | G.Skill RipJaws DDR4 3600 16GB | Samsung 970 Pro 512GB M.2 NVMe | (2) WD Blue 3TB Backups | MSI 2060 Super Gaming X | Meshify - C | SB X-Fi
- dicecca112
- Site Admin
- Posts: 5014
- Joined: Mon Mar 01, 2004 10:40 am
- Contact:
Coretemp does the exact same thing that TAT does but it works with newer AMDS and C2Ds
Last edited by dicecca112 on Sun Feb 25, 2007 2:03 pm, edited 1 time in total.
- Sparky
- Legit Extremist
- Posts: 727
- Joined: Fri Feb 16, 2007 4:47 pm
- Location: Huntington Beach, CA USA
Not sure about TAT and AMD cpu's but not willing to risk it.
Just removed, cleaned and reinstalled my heat sink & fan (Zalman 9500) along with some Arctic Silver 5. Still see difference in temps but at least it puts my mind at ease.
Just removed, cleaned and reinstalled my heat sink & fan (Zalman 9500) along with some Arctic Silver 5. Still see difference in temps but at least it puts my mind at ease.
MSI Meg X570 Unify | Ryzen 3900X | Noctuna NH-U12A | G.Skill RipJaws DDR4 3600 16GB | Samsung 970 Pro 512GB M.2 NVMe | (2) WD Blue 3TB Backups | MSI 2060 Super Gaming X | Meshify - C | SB X-Fi
- Sparky
- Legit Extremist
- Posts: 727
- Joined: Fri Feb 16, 2007 4:47 pm
- Location: Huntington Beach, CA USA
Core Center shows cpu usage and temp but I'm not sure about indiviual cpu temps. Still, I have always felt better knowing what my temps werre rather than assuming all was well.
MSI Meg X570 Unify | Ryzen 3900X | Noctuna NH-U12A | G.Skill RipJaws DDR4 3600 16GB | Samsung 970 Pro 512GB M.2 NVMe | (2) WD Blue 3TB Backups | MSI 2060 Super Gaming X | Meshify - C | SB X-Fi
- dicecca112
- Site Admin
- Posts: 5014
- Joined: Mon Mar 01, 2004 10:40 am
- Contact:
dicecca112,
It's all about the accuracy of the DTS in the die.
I noticed something was wrong with CoreTemp one day when I plugged a temp sensor in my digital multimeter and got 23ºC for ambient. CoreTemp was giving idle temps of 19/20ºC core#0/core#1 (E6600). Then I rigged the sensor on top of of the CPU waterblock right next to the lineout tube: DMM gives 27ºC.
Thus CoreTemp is off atleast by 8ºC.
But likely the offset in CoreTemp is >8ºC in reality because the top is likely not the warmest part of the waterblock.
The DMM is 1.5ºC accurate.
It's all about the accuracy of the DTS in the die.
I noticed something was wrong with CoreTemp one day when I plugged a temp sensor in my digital multimeter and got 23ºC for ambient. CoreTemp was giving idle temps of 19/20ºC core#0/core#1 (E6600). Then I rigged the sensor on top of of the CPU waterblock right next to the lineout tube: DMM gives 27ºC.
Thus CoreTemp is off atleast by 8ºC.
But likely the offset in CoreTemp is >8ºC in reality because the top is likely not the warmest part of the waterblock.
The DMM is 1.5ºC accurate.
Hmm, you may have received a CPU with dodgy on-die temp sensors but I doubt it.largon wrote:dicecca112,
It's all about the accuracy of the DTS in the die.
I noticed something was wrong with CoreTemp one day when I plugged a temp sensor in my digital multimeter and got 23ºC for ambient. CoreTemp was giving idle temps of 19/20ºC core#0/core#1 (E6600). Then I rigged the sensor on top of of the CPU waterblock right next to the lineout tube: DMM gives 27ºC.
Thus CoreTemp is off atleast by 8ºC.
But likely the offset in CoreTemp is >8ºC in reality because the top is likely not the warmest part of the waterblock.
The DMM is 1.5ºC accurate.
Also, CoreTemp may seem inaccurate for your CPU but that doesn't mean it's wrong for eveyone elses.
To the original poster, I do find it surprising that you're getting different temps at full load. From my experience, though they may idle at slightly different temperatures, the max load temps will be the same.
This is similar with the QX6700, which is two dual cores in the one CPU, where you can get differences between cores but generally core#0 and core#1 (CPU1) will have the same load temp and core#2 and core#3 (CPU2) will too. There may be a difference between CPU1 and CPU2 though, sometimes by 5C or more.
Maybe download Orthos and run the Small FFT test for 10 minutes and see if the temps end up being the same. If not you may need to make sure your HSF is applied with equal pressure.
- Sparky
- Legit Extremist
- Posts: 727
- Joined: Fri Feb 16, 2007 4:47 pm
- Location: Huntington Beach, CA USA
I just ran Stress Test v1.0 for 2 minutes and both cores still show 2-4 degrees C difference. Got close a couple of times, but then differed again. Not too worried about it because I've read more and see it appears to be normal.
MSI Meg X570 Unify | Ryzen 3900X | Noctuna NH-U12A | G.Skill RipJaws DDR4 3600 16GB | Samsung 970 Pro 512GB M.2 NVMe | (2) WD Blue 3TB Backups | MSI 2060 Super Gaming X | Meshify - C | SB X-Fi