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

How much thermal paste is needed for Mac Pro 1,1 CPUs?

PowerMac G4 MDD

PowerMac G4 MDD

macrumors 68000
Original poster
Jul 13, 2014
1,900
277
Hi everyone,

I completely disassembled my 1,1 Mac Pro (except for the motherboard) in order to clean it out, etc. I had to remove the CPU heat-sinks in order to remove the PSU...

So that means I had to re-apply thermal paste. I put a dot on each CPU (somewhat larger than a grain of rice each); however, now I am all paranoid and wondering if that was enough thermal paste to have applied.

Should I have spread it out evenly over the surface, or was putting a blob in the middle okay? I don't think my CPUs are running at an abnormal temperature; the room is a bit stuffy, and the readings seem as though they are similar to what they were before I took the Mac Pro apart. I just wanted to be sure, though.

-Thanks

P.S. I know that it may seem stupid that I am asking this after putting the entire thing back together, but, should I have any issues, I can easily take it apart again. I mean, I prefer NOT to go through that again, but I've done it before and am completely comfortable with it.
 
No pre-spread, that can introduce air bubbles. Rice grain size is about right. Cross the blob into an X shape for more corner coverage if you want.
 
Demigod Mac said:
No pre-spread, that can introduce air bubbles. Rice grain size is about right. Cross the blob into an X shape for more corner coverage if you want.
Click to expand...

Alright; yeah, I made it a little larger than a rice grain. I was worried that the temps seemed higher than before, but I haven't tested full idle yet at normal room temperatures. I cannot say how it is now because I am exporting a video in iMovie and have Facebook open. Currently, that puts the CPU A Proximity at 131F, and my Northbridge heat-sink at 165F. iMovie is using '280%' CPU, so that's obviously why.
 
I spread out an even layer to cover the whole pad area on each cpu. Had Handbrake running at 780% cpu load and my temps (core) reached 136*F, cooler than the old processors ran. Cpu B does tend to ran about 9 degrees warmer than cpu A when stressed.
 
I have worked on several 1,1 Macpros and found they were about the worse of any computer I've worked on for having trouble with the thermal paste. I would strongly recommend that you apply a coating to both cpus with a credit card, and then put a dab smack in the middle. Then install the heat sinks, turn it on and let it warm up, you don't even have to fully reassemble the memory cage if you test it on its side, you can put in a memory tray without the frame to test it temporarily.
Then disassemble and check the pattern, often one or both will not have spread well and only parts will have contacted. If both look like the paste was in full contact, put a little bit in the middle, and reinstall and everything will be great, if it connected poorly, put a bunch in the middle and try again. There is really very little harm other then wasting the paste in using slightly more then necessary, but there can be a lot of problems if you don't use enough, and the 1,1 macpro usually requires more then normal.
 
Thermal paste quality matters too. Recommend getting Arctic MX-4 or Gelid GC Extreme for best results.
 
dfritchie said:
I spread out an even layer to cover the whole pad area on each cpu. Had Handbrake running at 780% cpu load and my temps (core) reached 136*F, cooler than the old processors ran. Cpu B does tend to ran about 9 degrees warmer than cpu A when stressed.
Click to expand...

Then I guess I may need to apply more...
 
Bubalight said:
I have worked on several 1,1 Macpros and found they were about the worse of any computer I've worked on for having trouble with the thermal paste. I would strongly recommend that you apply a coating to both cpus with a credit card, and then put a dab smack in the middle. Then install the heat sinks, turn it on and let it warm up, you don't even have to fully reassemble the memory cage if you test it on its side, you can put in a memory tray without the frame to test it temporarily.
Then disassemble and check the pattern, often one or both will not have spread well and only parts will have contacted. If both look like the paste was in full contact, put a little bit in the middle, and reinstall and everything will be great, if it connected poorly, put a bunch in the middle and try again. There is really very little harm other then wasting the paste in using slightly more then necessary, but there can be a lot of problems if you don't use enough, and the 1,1 macpro usually requires more then normal.
Click to expand...

Yeah, it seems as though the Mac is running just a little bit hotter at idle than before, which really bothers me. I think I'll try adding some more thermal paste later.
 
dfritchie said:
I spread out an even layer to cover the whole pad area on each cpu. Had Handbrake running at 780% cpu load and my temps (core) reached 136*F, cooler than the old processors ran. Cpu B does tend to ran about 9 degrees warmer than cpu A when stressed.
Click to expand...

What kind of thermal paste did you use? Someone else here suggested Arctic, which I am thinking of getting. (BTW, do you recommend spreading a layer and then putting an extra blob/grain in the middle? Someone else here recommended that, as Mac Pro CPUs need more paste than usual. (Also, how many grams is necessary?)
 
I got my processors off eBay from Korea, I just use the thermal grease they sent. I put it on according to the Apple service manual which states " even layer 2 mm thick". So far all is good, actually runs cooler, idles at 120*F, stock ones idled at 131*F. All in all, best $80 I've spent on the old box.
 
Arctic MX-4 is a good price/performance paste (not to be confused with Arctic Silver, btw)
Best paste is Gelid GC-Extreme.

Some vids that might help:

The paste is only there to fill the microscopic gaps between the otherwise flat, tightly pressured metal surfaces, so you don't need a lot.

I'd recommend doing the rice grain, X or 5-dot method.
 
Demigod Mac said:
The paste is only there to fill the microscopic gaps between the otherwise flat, tightly pressured metal surfaces, so you don't need a lot.
Click to expand...

+++++

This is the reality that most don't understand.

The Delta T of metal is better then any paste that isn't just ground up metal. So, metal to metal will transfer MORE heat then a layer of paste between. You are ONLY filling the microscopic holes in the surface, not coating it. If you have an even 1mm coating you are greatly REDUCING the ability to cool your CPU.
 
dfritchie said:
I got my processors off eBay from Korea, I just use the thermal grease they sent. I put it on according to the Apple service manual which states " even layer 2 mm thick". So far all is good, actually runs cooler, idles at 120*F, stock ones idled at 131*F. All in all, best $80 I've spent on the old box.
Click to expand...

I thought you said that yours went up to 130 at load. Before I re-applied the paste, my Mac started at 85F and rested somewhere at 97-100F idle. Now, with the re-application, it seems to run at 104F idle at the coolest.
 
Demigod Mac said:
Arctic MX-4 is a good price/performance paste (not to be confused with Arctic Silver, btw)
Best paste is Gelid GC-Extreme.

Some vids that might help:

The paste is only there to fill the microscopic gaps between the otherwise flat, tightly pressured metal surfaces, so you don't need a lot.

I'd recommend doing the rice grain, X or 5-dot method.
Click to expand...

Thanks for that; I was almost considering spreading the thermal paste, but that evidently isn't the best idea. Maybe I just put a larger pea in the middle or something? I put a grain that was a little larger. I used generic silicon paste from China, so maybe that's a factor here?
 
Someone mentioned that 'IC Diamond' thermal paste helped his Mac Pro's temps a lot. He used Arctic Silver prior to that and didn't like his temps. I know you said not to confuse MX-4 with silver... but what is the difference between the two? Are you sure that MX-4 will allow for low temperatures? I'm wondering if I should skip Arctic altogether.

(BTW, my Mac Pro, before re-applying the thermal paste and all, ran at 96F-100F idle. I'm looking for similar temps this time. I think I'll go with the larger pea method for this.
 
A reputable brand name paste can make a noticeable difference over the cheap/generic stuff.

MX-4 and Silver 5 are from two different companies entirely despite both having 'Arctic' in them.
Silver 5 is the most famous paste; still good but no longer the king.
IC Diamond has good performance but is abrasive on the surfaces due to the particles, so I'd avoid it generally.

I'd recommend GC-Extreme or MX-4 having used both personally. Easy to apply, high performance, no-nonsense pastes. MX-4 is half as expensive but almost as good, but if you want the best, GC-Extreme is the one to get.

http://www.tomshardware.com/reviews/thermal-paste-performance-benchmark,3616-20.html has a roundup.

(though honestly, agonizing over picking a high quality paste is splitting hairs over a 1-2˚C difference) :p
 
Demigod Mac said:
A reputable brand name paste can make a noticeable difference over the cheap/generic stuff.

MX-4 and Silver 5 are from two different companies entirely despite both having 'Arctic' in them.
Silver 5 is the most famous paste; still good but no longer the king.
IC Diamond has good performance but is abrasive on the surfaces due to the particles, so I'd avoid it generally.

I'd recommend GC-Extreme or MX-4 having used both personally. Easy to apply, high performance, no-nonsense pastes. MX-4 is half as expensive but almost as good, but if you want the best, GC-Extreme is the one to get.

http://www.tomshardware.com/reviews/thermal-paste-performance-benchmark,3616-20.html has a roundup.

(though honestly, agonizing over picking a high quality paste is splitting hairs over a 1-2˚C difference) :p
Click to expand...

Well, I feel as if the cheap stuff isn't going to do the Mac Pro any good, and I am certain that the larger grain method did not cover as much of the CPU as I'd like... also, If you get to know me, things like this (in this case, a few degrees of temperature difference) drive me crazy. I'm already going to get rid of 1-2 of my HDDs and just get an SSD, as my frontal HDD is adding tons of noise to this Mac Pro. I was questioning the integrity of my system, as others were saying that their 1,1 Mac Pros were dead silent.

I think I'll just get the better thermal paste and try the blob/pea method and call it a day. It's annoying to disassembly a Mac Pro, but this will have been my fourth time. Once done once, it's not too big of a deal anymore. That being said, I'd like to not jinx myself here!
 
pastrychef said:
The Power Mac G4 MDD must have really drove you bonkers.
Click to expand...

Haha--I actually used one for a few months. It was annoyingly loud, but I do respect the machine. What's interesting, though, is that my Quicksilver is pretty much as loud as my MDD. I'm not sure what Quicksilvers other people have, but mine sure as heck isn't quiet when compared to my MDD. I'm not sure why my MDD didn't drive me crazy; I guess noise is getting to me even more by now, as I have lived off second-hand machines for some time now. (My last brand-new Mac--or computer in general--was purchased in 2004.) I used to be fine with my iMac G5, but it's getting to be a bit annoying now. As for the Mac Pro, it annoys me even more because it SHOULD be quiet but isn't. I found out that my primary HDD is creating most of the extra noise, so I'm getting an SSD for this soon (which was my intention in the first place). The GPU is driving me crazy, as it SHOULD be quiet. And, now, the temps are driving me crazy, even if they are only around 4 degrees F higher than usual. Gonna buy thermal paste and tear my Mac Pro apart AGAIN in order to fix that.

It's not too healthy to be so picky about every little thing, but perfectionism has its bonuses.
 
Wow. Your experiences with the MDD and Quicksilver are very, very different from mine. My Quicksilver was extremely quiet and my MDD was insanely noisy. I went through the trouble of replacing each and every single fan in the MDD in an attempt to quiet it down... But it remained my favorite Power Mac because it was the most powerful model that ran OS 9.

Ever since I started using Mac Pros, I've been spoiled by their silence. At one point, I had 130W 3.2GHz CPUs in my current 5,1 and it caused the booster fans to ramp up more frequently and sooner than I had been used to and eventually changed the CPUs to 95W X5670s which returned the booster fan behaviors back to normal.

I have also been careful to choose efficient graphics cards and pay attention to the types of coolers they use. Cool, quiet, and efficient outweigh utmost performance for me.
 
pastrychef said:
Wow. Your experiences with the MDD and Quicksilver are very, very different from mine. My Quicksilver was extremely quiet and my MDD was insanely noisy. I went through the trouble of replacing each and every single fan in the MDD in an attempt to quiet it down... But it remained my favorite Power Mac because it was the most powerful model that ran OS 9.

Ever since I started using Mac Pros, I've been spoiled by their silence. At one point, I had 130W 3.2GHz CPUs in my current 5,1 and it caused the booster fans to ramp up more frequently and sooner than I had been used to and eventually changed the CPUs to 95W X5670s which returned the booster fan behaviors back to normal.

I have also been careful to choose efficient graphics cards and pay attention to the types of coolers they use. Cool, quiet, and efficient outweigh utmost performance for me.
Click to expand...

I'm glad to be talking to someone who cares as much about quietness as I do! I feel like many people would recommend a card or something and say that it's quiet... but then you realize that their idea of 'quiet' is nothing like your own. Have you ever had experience with the Radeon 5770? I got a flashed one for my Mac Pro, but it's really loud. I think the fan is worn out (beyond repair; I tried oiling it), or maybe something else. As I may have mentioned to you, I found a dealer on eBay who has very clean, "quiet" 5770s that he sells. I'm almost thinking of selling mine and getting one from him. I'm looking for the fan-blower type again because, when they work, they're apparently silent. I sorta don't trust the ones with the larger fan blades.
 
Demigod Mac said:
Looks about right. Here are Amazon links.

http://www.amazon.com/ARCTIC-Performance-Compound-Interface-Material/dp/B0045JCFLY/
http://www.amazon.com/Gelid-Solutions-GC-Extreme-Compound-TC-GC-03-/dp/B002P5W4RU/

The GC-E will do about 1-3˚C cooler than the MX-4, by the way. Up to you if it's worth the extra 5 bucks (I would say yes)
Click to expand...

As much as I'd rather not spend that much on a tube of gunk, I also think that it will be worth-it. Spending a few extra bucks and having great results will cause me to not lose my head. Now that you've witnessed how I am, you should see how scarily neat my desk is. xD

Thanks for the links--I'll probably get the better one whenever I'm ready to work on my Mac Pro.
 
As an Amazon Associate we earn from qualifying purchases.
You must log in or register to reply here.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.
Top Bottom
Back