Adding Good Thermal Compound to Mac Mini

[Guest]

So today I decided that I have had enough of my Mac mini's fan ramping up to 4400 and even 5400 RPM over uploading videos, watching 1080p60 on YouTube, playing Porting Kit Windows games, etc.  I think MacOS Sierra had something to do with this, but I also suspected decreasing effectiveness on whatever thermal was applied to the CPU and GPU (soldered to the board - so much for even obscure upgradability unless you really know what you're doing).  I used my various bits (thanks Linus for your endless iFixit promotions - they helped me!) to get in that tiny case, flipped the board over, took off the heat syncs for the GPU and CPU, took off the old thermal, added MX-4 Arctic thermal (which was non-conductive), put it back together, and now what?  Well, we needed a stress test.  Not wanting to buy software just for making this post, I decided to go with the free method of asking Terminal to produce a lot of the word "yes" repeating over and over, on all threads.  I did that stress test for 5 minutes or so, and interestingly enough, it seems the new compound works — I never got passed 4400RPM, and when the fan got to that level, the CPU actually cooled from 78C to 73C, whereas it normally goes into the 80's when stressed.  This is fantastic.  I still have to test raw upload, RAM, and PortingKit tests, but I'll keep you posted.  Here are some relevant screenshots, linked on my Google Drive: check them out (if you're at all interested).

 

https://drive.google.com/file/d/0BwaQzL0vJPfWYnVBajdMQ1UwajQ/view?usp=sharing

https://drive.google.com/file/d/0BwaQzL0vJPfWaEdjdkR0Qzh3MUE/view?usp=sharing

https://drive.google.com/file/d/0BwaQzL0vJPfWZ1IwOVFOWk8tMzQ/view?usp=sharing

https://drive.google.com/file/d/0BwaQzL0vJPfWNnZRaTd2d1VFcjQ/view?usp=sharing

 

I hope this post is relevant.  If you want to ask questions about my methodology or want to suggest better or other ways I could stress test, let me know!  Unfortunately, I don't have stress test data for the before the new compound, but I can tell you, my CPU thermal throttled to 75% full usage in MONIT.

 

Edit: My Mac Mini Model is the Mac Mini 5,2 (Mid 2011); I added a Mercury 6G 240GB SSD and 8GB of DDR3 1333MHz SODDIM Crucial memory to it.  It's GPU, a Radeon 6630M, is soldered to the board.  I believe the CPU is a Sandy Bridge Core i5-2520M dual-core hyper-threaded processor.

[SCHISCHKA]

i have a 2009 mac mini with a noisy fan & a dead dvd drive. im looking to replace the heatsink entirely with something stupidly massive like a scythe ninja if i can get my hands on one

[SCHISCHKA]

2 hours ago, VPrime said:

I hope this post is relevant.  If you want to ask questions about my methodology or want to suggest better or other ways I could stress test, let me know!  Unfortunately, I don't have stress test data for the before the new compound, but I can tell you, my CPU thermal throttled to 75% full usage in MONIT.

i forgot to add this i use dd if=/dev/zero of=/dev/null on unix systems. i execute that in as many terminals as i have cpu cores and it puts the cpu straight to 100%

[Guest]

13 hours ago, SCHISCHKA said:

i forgot to add this i use dd if=/dev/zero of=/dev/null on unix systems. i execute that in as many terminals as i have cpu cores and it puts the cpu straight to 100%

Yeah I got that good 100% usage on it!  Tell me, is there any difference from the commands I used with the commands you used?

[SCHISCHKA]

5 hours ago, VPrime said:

Yeah I got that good 100% usage on it!  Tell me, is there any difference from the commands I used with the commands you used?

I don't think so. It's basically executing some x86 code with zero bottlenecking from the hard drives or anything else