I am an owner of a Mac Pro 6,1 (Late 2013). Yeah, I didn’t buy it - a job gave it to me. It’s a 6-core currently and I’m thinking of dropping a 12 core in it and using it till the thing falls apart. It fits my needs, it plays WOW Classic. Really don’t need anything else.
Just curious what people think is the better way to spread TIM on the Xeon in this machine? I know Apple’s heatsink design in this thing is …unconventional. Usually I’m a fan of the small bead method but after looking online - it looks like maybe spreading it across the IHS would make more sense given the pressure of the plate.
Feel free to tell me I’m overthinking it.
(Yes I know you all hate the MP6,1 don’t tell me - I didn’t spend money on it since the day I got it so I really don’t care)
I have one, and I bought it when it first came out. I don’t hate it, but I hate Apple for abandoning it.
However… I would not spend money upgrading it. By the time you upgraded the CPU, and then the SSD, you’d have a lot of the money for buying a much better PC. Yeah, I know, I hate that idea, but Apple has finally abandoned the mid-range user. I don’t want an iMac with a Vega 48 GPU, and the iMacPro (and the upcoming MacPro) are stupidly expensive. So there’s not really a good mid-range Apple option anymore.
To answer your question - you should spread a thin layer of TIM across the surface of the IHS. I concur that the Mac Pro 6,1 is not ideal these days - however if it works for your work flow then go for it. I personally am using an iMac Pro and a MacBook Pro 16" for my work and love both.
If the CPUs on these Mac Pro’s use a heat spreader and the heatsink is mounted with a fair amount of pressure, you might be able to use one of the thermal pad solutions that’s come out in the past couple of years like Innovation Cooling’s graphite pads (electrically conductive) and Thermal Grizzly’s Carbonaut (electrically nonconductive). They perform just a few degrees worse than a decent paste and take a lot of the guesswork out of reapplying TIM. They never dry out so you won’t need to reapply them, and when you’re done with the Mac Pro you can remove the pad and reuse it elsewhere without degradation.
Unless you meet the two criteria below, don’t use a graphite thermal pad:
- You can get the motherboard perfectly flat so that the pad doesn’t slide while you prep the heatsink. This is important, as the pad does not have any adhesive and will move if there’s any slant to the motherboard.
- The CPU’s IHS and heatsink surfaces are both smooth as glass with no etching at all. This is the killer here. These pads do not form into micro-grooves like thermal grease and regular thermal pads do. If both surfaces aren’t perfectly smooth, you’re going to be better off with thermal grease, preferrably non-conductive given the nature of the MP6,1’s vertical configuration.
I would recommend something along the lines of Noctua’s NT-H2 for your TIM as it is non-conductive and has an exceedingly long usable lifespan. My NT-H1 application is six years old now and is still going just as strong as it did on day one (though I plan on reapplying fresh NT-H2 when I swap this rig’s innards into the HAF XB EVO case I got so I can put my new components into this case.
In all honesty, I’d stick with thermal grease/paste. That extra bit of cooling can mean a lot with a heat sensitive setup like the MP6,1.
That second point is interesting, would the etching on the surface of AMD CPUs be a problem? A friend of mine did a build with a 3700X and the IC pad and so far it seems to work nicely.
Also just out of curiosity, how would you rate the Noctua pastes against MX-4? The latter is all I’ve ever used and it has yet to pose any kind of problem, even after several years.
Let’s just say that if there is etching, you’re better off just using regulra thermal paste/grease since that will fill the pits whereas a graphite pad will not. Air gaps make for hotter machines. That’s why I said don’t use them unless both the IHS and heatsink have mirror finishes on them. Both.
MX-4 is just fine. If you have some that is less than two years old (unopened), it’s good for about another year. Past two years I wouldn’t trust it all that much. But that’s just shelf life. Usable on-die life can be five or six years easily. Eventually the TIM does need to be replaced though as its thermal characteristics have deteriorated with both aging and constant flex/contract cycles. You’d be surprised what TIM from a decade ago would likely look like if you took off the heatsinks. Many would easily be closing in on being as hard as clay (and virtually useless as a thermal interchange material anymore).
Once that tube is opened though, you have two years max before it becomes worthless. Once the vaccuum seal has been broken on the inside of the tube, hardening begins. It’s very slow, but it does happen. Caulk is the same way.