3

Before I just rush into ripping apart my machine & putting new thermal paste on the CPU heatsinks, I thought I'd try running this past Ask Different.

I have a 'Franken-Mac' Pro 4,1 with the full 5,1 upgrade.
It now has dual 6-core 3.46GHz Xeons - X5690
64GB RAM

From comments
'Franken-Mac' = Non-stock model. Mac Pro 4,1, upgraded to higher 5,1 spec than was available when new. All Apple, but there was never a 3.46 model originally.

I have no idea what was in it before, as I bought it ready-upgraded.

What worries me is that the 'main' temperatures reported by iStat Menus - which on this machine are the CPU A & B Tdiodes [not that I really know what they are] - are consistently out by about 15°C
Under minimal load, this comes out as approx CPU A at 60°C & CPU B at only 45°C
If I put the machine under load, I've seen that lift to perhaps 75°C & 60°C
The booster fans always run at the roughly same speed as each other, whatever that speed might be, increasing with load of course.

enter image description here

BTW, it's not always exactly 15° but around that value.

I have an older Mac Pro 3,1 which, though iStat measures a different sensor for the 'main' temperature [CPU A & B Heatsink], is always within a degree, right now 35/36°C

enter image description here

I'm not worried about the apparent temperature difference between the two machines, as not the same thing is being measured.

So, my question is...
Does anyone know if this difference between CPU A & B temperatures is normal, or if I need to be looking at the thermal paste etc?

The airflow is clean. I do it regularly anyway & did it just before I posted this question.

I've never stripped a Mac right down to the CPUs before [though I've done dozens of generic PCs over the years] so I wanted to get a second opinion.

A small side question - does anyone know which is A & B, physically? Is A to the front, B to the rear? I know they're labelled on the PCB, but it's not visible until you strip them & I can't find a picture of sufficient clarity.

Solution to the side question - CPU A is to the rear of the machine, so the air hitting it will have already passed over CPU B's heat-sink.

enter image description here

  • 'Franken-Mac' ? – user3439894 Jun 23 '16 at 13:13
  • Non-stock. 4,1, upgraded to higher 5,1 spec than was ever available when new. All Apple, but there was never a 3.46 model from new – Tetsujin Jun 23 '16 at 13:18
  • The TCASE temp for your CPU is 78.5°C, so as long as it's not going over that, I wouldn't muck with it. That said, if you can easily reapply the thermal paste, then go for it. – user3439894 Jun 23 '16 at 13:28
  • I had a Dell Precision workstation back in the day, all factory equipped, that would do this same thing. From what I found out, this was normal because one CPU had "management overhead" and required more cycles. I can't seem to find the material that references it, so I don't want to post as an answer just yet. – Allan Jun 23 '16 at 14:16
2

I upgraded my 4,1 Mac Pro to a pair of X5690s and am seeing exactly the same Tdiode temperatures as you - CPU A around 61°C and CPU B around 46°C at idle.

It seems to be normal owing to the inline airflow and discrepancies in heatsink size. As others have said, as long as the Tcase max temp is not exceeded on either CPU, I think we're fine and can rest safe in the knowledge that it's not due to bad/inconsistent thermal paste application.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .