Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

buster84

macrumors 6502
Oct 7, 2013
428
156
Ok, so I went a little crazy since I had a really bad day and then coming home to this was the last drop, well, I just said F*#$ it and did the following:


Grabbed the old board and installed the old CPU's onto it, just like that with the already on thermal paste on cpu and heatsink and i did not count how many turns i gave to the cpu and i just went with it until i was like maybe one more will break this sh@*%. Plugged the old board into the case and WTF it turned on! I got kernel panics but it TURNED ON!!!

After I saw that the old board was giving kernel panics non stop i thought, well, this has to be related to the bent pins so lets put the new board and do the same, just fuc@# it.

So i did, and IT TURNED ON!! But without kernel panics!! So now i have it working but I am afraid that since I did not reapply the thermal paste again and kept on just swapping and trying, swapping and trying this and that, the temp is going to get weird.
I started monitoring the CPU temp with a tool I have and I see I have a 9 to 10 degree difference between the Diodes/Heatsink/Core PCECI between CPU A & B. Not sure if this is normal of if both CPUs should be the same.

Nice! glad you got it working so far. So it sounds like your main issue was not tightening it down good enough. If you got kernal pannics its possible that you still needed to tightening it down a tiny bit more. If i was you, id test the other processors in the first trey and see if you can get it working without any problems. Rebend an bent pins back into place, but do it cautiously, to much it'll rip right out. I'd suggest a tooth pick that way your just pushing instead of tweezers which will pull.

Also a side note/warning. When it comes to these mac pros and the processors, never touch them when your having a bad day. These installs need 100% of your attention.
 
  • Like
Reactions: Sharky II

h9826790

macrumors P6
Apr 3, 2014
16,656
8,584
Hong Kong
Ok, so I went a little crazy since I had a really bad day and then coming home to this was the last drop, well, I just said F*#$ it and did the following:


Grabbed the old board and installed the old CPU's onto it, just like that with the already on thermal paste on cpu and heatsink and i did not count how many turns i gave to the cpu and i just went with it until i was like maybe one more will break this sh@*%. Plugged the old board into the case and WTF it turned on! I got kernel panics but it TURNED ON!!!

After I saw that the old board was giving kernel panics non stop i thought, well, this has to be related to the bent pins so lets put the new board and do the same, just fuc@# it.

So i did, and IT TURNED ON!! But without kernel panics!! So now i have it working but I am afraid that since I did not reapply the thermal paste again and kept on just swapping and trying, swapping and trying this and that, the temp is going to get weird.
I started monitoring the CPU temp with a tool I have and I see I have a 9 to 10 degree difference between the Diodes/Heatsink/Core PCECI between CPU A & B. Not sure if this is normal of if both CPUs should be the same.
CfBwXps.png


Benchmark:

3d1lQgI.png

All temperature looks good to me.

With extremely good thermal paste application (may be with some luck as well). And the Mac is at true idle (but not very very low usage). The temperature difference between CPU and it's heatsink can reach 1C. However, in general 3-5C is normal.

That PCECI is not a good reference, you can simply ignore that.

10C difference between CPU A and B is also normal. That's because heat sink A is at down stream of the airflow.

However, this cannot shows if the thermal paste is really working well, unless you capture the fan RPM and CPU temperature's during stress test. In my personal experience, once you start the stress test, you can know the answer straight away, no need to wait for the temperature stabilised. Because on a poor cooling system (poor thermal paste application / dusty heatsink / broken fan / etc), the temperature will rise very quickly, easy to reach 75C or above in few seconds, and the fan will spin up quickly. But on a good system, the temperature may go up to 6xC range in few seconds, but then take much longer time to slowly reach the stabilised temperatures / fan RPM.
 
Last edited:

databaze

macrumors member
Original poster
Nov 21, 2011
53
8
All temperature looks good to me.

With extremely good thermal paste application (may be with some luck as well). And the Mac is at true idle (but not very very low usage). The temperature difference between CPU and it's heatsink can reach 1C. However, in general 3-5C is normal.

That PCECI is not a good reference, you can simply ignore that.

10C difference between CPU A and B is also normal. That's because heat sink A is at down stream of the airflow.

However, this cannot shows if the thermal paste is really working well, unless you capture the fan RPM and CPU temperature's during stress test. In my personal experience, once you start the stress test, you can know the answer straight away, no need to wait for the temperature stabilised. Because on a poor cooling system (poor thermal paste application / dusty heatsink / broken / etc), the temperature will rise very quickly, easy to reach 75C or above in few seconds, and the fan will spin up quickly. But on a good system, the temperature may go up to 6xC range in few seconds, but then take much longer time to slowly reach the stabilised temperatures / fan RPM.


I'll be monitoring the temps for the next few days, hopefully this is the end of all issues for me and I can finally use it.

Thanks everyone for all your recommendations and help, I really appreciate it, this is a great community to be a part of.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.