10C difference between same CPU

[skaughtz]

I have two computers with almost identical setups.  Silverstone Grandia GD09B case, i5-3570k overclocked to 4.2 Ghz with a -0.45V offset, Cryorig m9 Plus cooler.  CPU A is on an Asrock Z75 Pro3 with a 50% LLC (options are 0, 50, or 100%), and CPU B is on an Asus P8Z77-V LX with a 25% LLC (options are greater).  Both boards are essentially operating under identical power and fan settings.

 

No matter how many times I reapply thermal grease or reseat the cooler, CPU A runs about 10C cooler than CPU B (Cores 1/2/3/4 under P95 = 56C/64C/60C/63C  and 69C/73C/76C/72C, respectively).  Is there a variable that I have overlooked involving the motherboards, or could there really just be that big of a difference in the silicon of the CPUs that one would run that much hotter than the other?

 

 

[Fasauceome]

Have you used software to confirm the operating voltage?

[ddennis002]

2 minutes ago, skaughtz said:

I have two computers with almost identical setups.  Silverstone Grandia GD09B case, i5-3570k overclocked to 4.2 Ghz with a -0.45V offset, Cryorig m9 Plus cooler.  CPU A is on an Asrock Z75 Pro3 with a 50% LLC (options are 0, 50, or 100%), and CPU B is on an Asus P8Z77-V LX with a 25% LLC (options are greater).  Both boards are essentially operating under identical power and fan settings.

 

No matter how many times I reapply thermal grease or reseat the cooler, CPU A runs about 10C cooler than CPU B (Cores 1/2/3/4 under P95 = 56C/64C/60C/63C  and 69C/73C/76C/72C, respectively).  Is there a variable that I have overlooked involving the motherboards, or could there really just be that big of a difference in the silicon of the CPUs that one would run that much hotter than the other?

 

 

The Thermal paste between die and IHS could be a factor as well.

[skaughtz]

12 minutes ago, fasauceome said:

Have you used software to confirm the operating voltage?

I can't believe I overlooked the obvious.  Checking back on the screenshots I took of Core Temp the cooler CPU is reading as 67.2W under full load while the hotter one is reading as 79.9W under full load.  There has to be a voltage difference between the two as I imagine the amps have to be near identical.

 

I assumed the boards would just offset from the same stock voltage value.  Now I have to figure out why that Asus board is offsetting a higher voltage than the other and how to lower it.

[bowrilla]

13 minutes ago, skaughtz said:

I can't believe I overlooked the obvious.  Checking back on the screenshots I took of Core Temp the cooler CPU is reading as 67.2W under full load while the hotter one is reading as 79.9W under full load.  There has to be a voltage difference between the two as I imagine the amps have to be near identical.

 

I assumed the boards would just offset from the same stock voltage value.  Now I have to figure out why that Asus board is offsetting a higher voltage than the other and how to lower it.

No, not necessarily the voltages. Most boards have an option to remove the current limits. This way, the CPU can draw more current resulting obviously to higher temps.

 

You know, you could switch the CPUs between systems and verify that it's in deed the board and not the difference between CPUs. You might need to readjust your OC settings though.

[skaughtz]

17 minutes ago, bowrilla said:

No, not necessarily the voltages. Most boards have an option to remove the current limits. This way, the CPU can draw more current resulting obviously to higher temps.

 

You know, you could switch the CPUs between systems and verify that it's in deed the board and not the difference between CPUs. You might need to readjust your OC settings though.

If I can't track down the cause on the board I might do that... but I just got them reassembled all nice with the cables bound and tucked away... 

 

Regardless, I am curious if it is plausible that it could just be a difference in the CPUs themselves.  A couple degrees difference wouldn't surprise me, but there is an 11.75C average core temperature difference here.

[bowrilla]

14 minutes ago, skaughtz said:

If I can't track down the cause on the board I might do that... but I just got them reassembled all nice with the cables bound and tucked away... 

 

Regardless, I am curious if it is plausible that it could just be a difference in the CPUs themselves.  A couple degrees difference wouldn't surprise me, but there is an 11.75C average core temperature difference here.

It is possible. A well overclockable chip with decent tim Vs bad overclocker and bad tim. Add in some variances due to boards and different VRMs...

 

Edit: just swap the chips. Then you'll know if it's the chips or the boards.

[LogicalDrm]

1 hour ago, skaughtz said:

I can't believe I overlooked the obvious.  Checking back on the screenshots I took of Core Temp the cooler CPU is reading as 67.2W under full load while the hotter one is reading as 79.9W under full load.  There has to be a voltage difference between the two as I imagine the amps have to be near identical.

 

I assumed the boards would just offset from the same stock voltage value.  Now I have to figure out why that Asus board is offsetting a higher voltage than the other and how to lower it.

LLC setting might be a thing. That differs between the boards (check out some OC3D motherboard reviews to get experts  opinion). But like said, only thing to see for sure is to use same CPU on both mobos and do the testing. Like a real reviewer. I've watched enough of TTL's testing to know how much variation you can get from two boards, and even between two early BIOS versions.