[nikhsub1]

Let me start by saying I just bought a brand new, less than a week ago, (from the Apple Store) 17" Santa Rosa MBP with the hi res screen (matte), 160GB 7200 RPM HD, 2.4Ghz CPU and the 256MB 8600 GT. For some more background I am a system administrator and a PC tech so I know what I'm doing with computers (most of the time anyway ). Anyhow, after getting my MBP and installing Vista on it via bootcamp and thoroughly enjoying the dual booting, I notice the thing runs hot. Ok, I do some google'ing and find the various thermal paste horror stories floating around. In windows (I've been spending more time in windows since I got it) I notice that it is IDLING at ~ 53-60C!!! I think perhaps this is not right, something may be wrong with the temp monitor software I use? Well, I use several, one of which is Intel softare called Intel Thermal Analysis Tool that monitors the DTS (Digital Thermal Sensor) in each core, it also will load the pants out of an Intel CPU - ~ 20% higher load than Intel's specified TDC. So for living with the heat for almost a week I decide I must bust the machine open, remove the logic board (mother board for you non apple folk) and see what kind of thermal paste job is in there.

Let me say that removing the logic board from these machines IS NOT FOR THE FAINT OF HEART!!! It scared the #@#$%$ out of me a few times! I took my time and marked where every last screw went and every cable, even the yellow tape. Once I had the board out and turned it over (the CPU, NB and GPU face DOWN so they can't be seen unless you remove the logic board) the sight was HORRIFIC!!! In my professional opinion they used about 40x too much paste, yes I said FORTY TIMES! All thermal paste is supposed to do is fill the micro groves and pits in the heatsink so VERY LITTLE IS NEEDED! Too much is worse than none. Anyhow, I'm sure you want pics so here they are. Oh, and my idle and LOAD temps dropped by ALMOST 20C!!!

Shame on whoever is assembling these machines as the average user will NEVER fix this and would perhaps think 'this is just how hot it is supposed to run'. I would demand Apple fix this, I would show them pictures etc, etc.

Oh the HORROR!




Now the heat pipe cooling system before:


Here is a CLEAN logic board - From left to right: GPU, North Bridge, CPU:


Proper paste applied - I have some special diamond dust thermal paste that is VERY durable so I used it - I have seen others apply the paste to the chips AND to the heat sinks, THIS IS BAD, VERY BAD!!! ONLY APPLY TO CHIPS.


And finally the cleaned heat pipe cooler:

[mduell]

Originally Posted by nikhsub1 

Shame on whoever is assembling these machines as the average user will NEVER fix this and would perhaps think 'this is just how hot it is supposed to run'.

Per the leaked assembly manual, the folks assembling them are following the directions they have been provided, and are presumably required to follow.

[MacosNerd]

Wow, I thought this thermal paste issue was a thing of the paste. There were quite a number posts in various forums regarding too much thermal paste on the MBP and MB. Kind of makes think twice about getting one.

I'm not about to take the laptop apart, while I'm not faint of heart, for me the prospect of voiding the warranty (if something broke) looms too large.

Its great that you're able to see such a huge decrease in temps.

[nikhsub1]

Everyone thinks this must void the warranty - to them I say how? There is no evidence that anything was opened on the notebook whatsoever. There were not stickers that said 'open me and your warranty is void'. I have a perfectly fine warranty TIA.

[MacosNerd]

Originally Posted by nikhsub1 

Everyone thinks this must void the warranty - to them I say how? There is no evidence that anything was opened on the notebook whatsoever. There were not stickers that said 'open me and your warranty is void'. I have a perfectly fine warranty TIA.

I didn't say opening it up voided the warranty but if you damaged the computer while working in it, that will void the warranty. The warranty is specific in that if a non authorized person works on the computer and breaks it, then apple is not going to honor that.

While there is no seals to break, if you cut a wire, or snap a pin off then that will be pretty evident and you'd be SOL

[nikhsub1]

Originally Posted by MacosNerd 

I didn't say opening it up voided the warranty but if you damaged the computer while working in it, that will void the warranty. The warranty is specific in that if a non authorized person works on the computer and breaks it, then apple is not going to honor that.

While there is no seals to break, if you cut a wire, or snap a pin off then that will be pretty evident and you'd be SOL

Ahh, understood.

[P]

Not again. Makes me wonder if I should pick apart my old G5 and check just in case.

[Tomchu]

Just to clear up a misconception ...

Thermal paste is all about the conduction of heat away from the chips to the heat pipes/case. If you do a proper application of high-quality paste, sure ... your chip temperatures will drop, but that's because you're conducting more heat away from them. And where's it going? To the case.

It seems almost counter-productive to perform this operation if your concern is a computer that is physically too hot for comfortable use, IMO.

[P]

Originally Posted by Tomchu 

It seems almost counter-productive to perform this operation if your concern is a computer that is physically too hot for comfortable use, IMO.

Not really. The amount of heat produced is essentially constant with temperature (it increases slightly with temperature, because of resistance, but let's keep it simple). That heat has to go somewhere - into the heatpipes where the fans can cool it, or into the case. In essence, the temperature of the case is a direct function of the interior temperature. The hotter the heatpipes and the attached sinks, the more heat is transferred out by the fans. The bigger the cooling effect of the fans, the cooler the interior of the case and the cooler the outside of the case.

[MacosNerd]

My concern is that I'm just about ready to pick on up. I just sold some stuff on ebay to help offset the cost and now I start seeing these types of posts.
thermal paste, uneven screens, 6-bit screens. While this does not appear that every machine is afflicted with this, its a troubling sign, however I'm wondering if I'm over analyzing it.

Of course this makes applecare all the more important especially with the heat issue, part failure may not show up in the first year but excessive heat will shorten component's life span

[Geofries]

See I'm in the same boat as you guys, ready to make the jump, but just seeing WAY too many posts about too many problems.

If I'm going to spend 3k, or near 3k, my machine needs to be absolutely perfect. Anything less is unacceptible.

[PaperNotes]

Seeing as the temperature drops by up to 20C with proper applied paste, are there jumper settings or any way to overclock the CPUs? No problem for me if the temperature goes back up 20C as long as the system can run much faster.

[MacosNerd]

I've been doing unscientific/unofficial research on this heat issue lately and so far most people are not reporting temps in the 60c range, though a few such as the OP are. I'm not sure if its related to specific place where the laptop was assembled or if a particular lot had this problem. Kind of disconcerting.

[MacosNerd]

I'm going to show my ignorance on the thermal paste issue so bear with me

Now the desired application of the paste is for a razor thin layer to fill in the imperfections on the core so the heat sink and cpu have a tight and complete seal to allow heat to radiate out of the cpu into the heat sink.

Since the the heat sink is pretty much bolted tight to the cpu thus squeezing excess thermal paste out what is the harm in applying too much. Wouldn't the end result be the same, a thin layer of paste between the cpu and heat sink?

[PaperNotes]

Originally Posted by MacosNerd 

Since the the heat sink is pretty much bolted tight to the cpu thus squeezing excess thermal paste out what is the harm in applying too much. Wouldn't the end result be the same, a thin layer of paste between the cpu and heat sink?

Excess paste spilling out on to the area around the CPU traps heat underneath it. Paste should be just enough to let heat rise not trap heat.

[MacosNerd]

Gotcha, but am I correct in assuming that heat will travel the path of least resistance and thus flow out of the heat sink if it cannot exit the side?

[Tomchu]

There is no heat being generated by the packaging of the chips -- just their cores. In this case, the cores are making contact with the heat transfer mechanism. I see no problem.

There is no heat being "trapped".

[soon2bmac]

I have the exact same laptop that I picked up 4 days ago and it runs at a steady 50C. Anyone else just pick one up with any other readings?

[nikhsub1]

Originally Posted by PaperNotes 

Excess paste spilling out on to the area around the CPU traps heat underneath it. Paste should be just enough to let heat rise not trap heat.

True to a point - it depends on mount PRESSURE between heat sink and die... if it is not so much, excess paste may not all be extruded - as was the case that I saw. There was WAY too much paste between sink and die. I will say that now that some time has passed and I've allowed the machine to run the gains (or temp loss) is not as great as originally thought. I see about a 9-12C drop in idle temps and about 16c load. Still very respectable but not the 20 I originally reported. Thermal management is crucial in laptops, every little bit helps. The application of paste in my computer was a disgrace and apple should be ashamed.

I do NOT recommend people removing mother boards to check this - if you think there is a problem take it back to apple. If your machine is idling close to or above 60c in OS X (i use smcfancontrol and istat pro) then I would bet you have way too much paste. Apple should be fixing this but it is a double edge sword. They will look at you like you are NUTS. They will wonder how the hell you think there is too much thermal paste on the CPU, NB and GPU. I guess you could show them pictures but I don't know that it would even help...

[mduell]

Originally Posted by Tomchu 

Just to clear up a misconception ...

Thermal paste is all about the conduction of heat away from the chips to the heat pipes/case. If you do a proper application of high-quality paste, sure ... your chip temperatures will drop, but that's because you're conducting more heat away from them. And where's it going? To the case.

It seems almost counter-productive to perform this operation if your concern is a computer that is physically too hot for comfortable use, IMO.

Please read a thermodynamics textbook before posting this incorrect FUD again.

[MacosNerd]

Originally Posted by mduell 

Please read a thermodynamics textbook before posting this incorrect FUD again.

Please explain?

Tomchu seems to be reaching on his assessment of using thermal paste because it will make the case hotter but I'm not sure what your point is.

[PaperNotes]

What about overclocking the CPU?
Which socket does this Merom use? Is it Socket P?
Is the CPU removable for future upgrades?

[ghporter]

Originally Posted by Tomchu 

Just to clear up a misconception ...

Thermal paste is all about the conduction of heat away from the chips to the heat pipes/case. If you do a proper application of high-quality paste, sure ... your chip temperatures will drop, but that's because you're conducting more heat away from them. And where's it going? To the case.

It seems almost counter-productive to perform this operation if your concern is a computer that is physically too hot for comfortable use, IMO.

Originally Posted by mduell 

Please read a thermodynamics textbook before posting this incorrect FUD again.

I am way confused by what Tomchu said, but I think I agree that he's not getting the basic idea of thermodynamics.

In a MacBook Pro, the case is supposed to receive the heat produced by the active components inside. It then radiates away the heat. Thermodynamics states, in brief and simplified terms, that heat energy flows "downhill" from a higher temperature system to a lower temperature system. Thus the heat in the CPUs and other chips is conducted to the cooler case and thence to the cooler ambient environment. Conducting more of the heat from the CPUs to the case WILL reduce the CPU temperatures. This is a Good Thing® and part of the cooling design of the computer. Other routes for heat to escape include the heat exchanger that has those lovely, almost inaudible fans blowing ambient air across it, thus radiating more heat to the external environment.

Finally, if the CPUs are, through a thin layer of thermal compound, in actual contact with the heat conductor assembly, then who cares that there is more compound than absolutely necessary? It's not reducing the contact between the CPUs and the heat conductor, and it's not adding noticeable weight. The compound won't leak out and mess up your clothes. What's the gripe?

[kman79]

I will be purchasing my first Apple Notebook tomorrow morning. I've been a PC user since I can remember, and I have been building my own PCs as well as PCs for other people for many years. Although my expertise is with Health Care, I do know that too much thermal paste is not a good thing. There is a very significant difference in Temperatures between applying the proper amount of thermal paste and applying way too much thermal paste. I know this from experience with building multiple PCs and overclocking ventures. It is very disappointing to see such poorly applied thermal paste with a notebook costing as much as a MBP. I guess in an effort to pump out as many Notebooks/Laptops as they can, there seems to be a decrease in the amount of time spent on properly applying thermal paste.

nikhsub1: Any chance in you writing up a DIY guide to dissasembly of the MBP?

[PaperNotes]

Originally Posted by ghporter 

Finally, if the CPUs are, through a thin layer of thermal compound, in actual contact with the heat conductor assembly, then who cares that there is more compound than absolutely necessary?

The thermal paste that gets squeezed out on to the rest of the CPU package traps heat under it and within the paste which not only increases CPU temperature but also decreases the lifetime of your processor.

This story is spreading. If you google "too much thermal paste" the MacBook Pro complaint has replaced all the top hits from the years since Google started indexing the net.

� Thermal grease theory | The Apple Core | ZDNet.com
Thermal Greasy: Apple Sics Lawyers on Something Awful - Gizmodo
Be cool with your Macbook Pro
MacDevCenter.com -- MacBook Pro: The Thermal Paste Question

[VValdo]

Originally Posted by PaperNotes 

This story is spreading. If you google "too much thermal paste" the MacBook Pro complaint has replaced all the top hits from the years since Google started indexing the net.

� Thermal grease theory | The Apple Core | ZDNet.com
Thermal Greasy: Apple Sics Lawyers on Something Awful - Gizmodo
Be cool with your Macbook Pro
MacDevCenter.com -- MacBook Pro: The Thermal Paste Question

Huh? these links are all describing the thermal paste issue from May 2006.

There are a new batch of threads though besides this one:

Apple - Support - Discussions - High CPU temp on new SR MBPs ...
Brand NEW 17" Santa Rosa MBP Thermal Paste Nightmare! - Notebook Forums and Laptop Discussion
Re-applying thermal paste on the Macbook - Notebook Forums and Laptop Discussion

W

[mduell]

Originally Posted by MacosNerd 

Please explain?

Tomchu seems to be reaching on his assessment of using thermal paste because it will make the case hotter but I'm not sure what your point is.

The amount of heat being moved (per time interval) is fixed*; it's whatever the CPU is putting out (say 20W). Moving heat is accomplished by temperature differences. Before and after changing the thermal paste, you're moving the same amount of heat from the case to the air, so the case temperature difference (above ambient air temperature) will stay the same. You're moving the same amount of heat from the heatsink to the case, so the heatsink temperature (above the case temperature) will stay the same. You're moving the same amount of heat from the chip to the heatsink, but you're doing it more efficiently because you now have metal-on-metal contact instead of thermal paste (which isn't that good at conducting heat, but it's better than air gaps) to go through, so the temperature difference between the heatsink and the chip is reduced by decreasing the temperature of the chip.

* It actually slightly increases as chip temperature goes up, and assuming constant load (obviously it changed with CPU load, but that's unrelated to re-doing your thermal paste).

Originally Posted by ghporter 

Finally, if the CPUs are, through a thin layer of thermal compound, in actual contact with the heat conductor assembly, then who cares that there is more compound than absolutely necessary? It's not reducing the contact between the CPUs and the heat conductor, and it's not adding noticeable weight. The compound won't leak out and mess up your clothes. What's the gripe?

Thermal paste is a better heat conductor than an air gap, but it's worse than metal on metal contact.

[MacosNerd]

^ Gotcha that makes sense.

[tmoney468]

I think mduell explained it well. The main problem with too much thermal paste is that is it prevents good metal-to-metal contact between the heatsinks and the CPU, thus raising the temperature of the CPU. All thermal paste is supposed to do is fill in the microscopic gaps in each surface (CPU and Heatsink) to create a good, clean, smooth contact.

[VValdo]

Originally Posted by VValdo 

There are a new batch of threads though besides this one:

Apple - Support - Discussions - High CPU temp on new SR MBPs ...
Brand NEW 17" Santa Rosa MBP Thermal Paste Nightmare! - Notebook Forums and Laptop Discussion
Re-applying thermal paste on the Macbook - Notebook Forums and Laptop Discussion

W

And to reply to myself, it's been dugg...
Digg - New MacBook Pros Still Overheating Due to Misapplied Thermal Grease
New MacBook Pros still overheating due to misapplied thermal grease? - Engadget

W

[ghporter]

Originally Posted by PaperNotes 

The thermal paste that gets squeezed out on to the rest of the CPU package traps heat under it and within the paste which not only increases CPU temperature but also decreases the lifetime of your processor.

"Traps heat under it?" How? Thermal paste is designed to TRANSMIT heat, not block it. The most basic thermal paste is zinc oxide in a silicone grease base, which does a pretty good job of transferring heat; current pastes are exceptionally good at that. How can a thermally conductive material "trap" heat? At worst, too much paste spread across the various heat pipe components could make it less efficient at moving heat from ONE source or another, but not enough to impact performance. I will research the links you posted, but I can not yet understand how too much heat conductor can "trap heat."

Originally Posted by mduell 

Thermal paste is a better heat conductor than an air gap, but it's worse than metal on metal contact.

Properly applied, with the heat sink properly attached, there is almost no paste between the chip and the heat sink-generally only the slightest film of the paste's carrier medium. This tends to equalize the heat transfer across the top of the chip. I would say "not as efficient as a perfect metal-to-metal junction" rather than "worse than metal on metal" because you can't get anywhere near a "perfect" junction in production.

I should point out that on my MBP, my fans are always running in the 1000RPM range (when I look at their speeds). The temperatures as reported by the latest version of iStat are consistent with heat being produced in the CPUs and being dissipated through the cooling system appropriately.

[indigoimac]

Ok, well this is interesting. Anyone crack open the latest revision 15"? I'm not about o but mines running at that same 59-60C idle in OSX, it'll kick up to close to 70-75 under load and maybe, just maybe the fans'll kick on. The Nvidia GPU hangs at about 10C hotter than the CPU.

My C2D iMac idles around 45C, I know it's a way larger enclosure, but it's still concerning that it's a 15C difference.

[PaperNotes]

Originally Posted by ghporter 

"Traps heat under it?" How? Thermal paste is designed to TRANSMIT heat, not block it. The most basic thermal paste is zinc oxide in a silicone grease base, which does a pretty good job of transferring heat; current pastes are exceptionally good at that. How can a thermally conductive material "trap" heat?

Heat from a CPU doesn't just radiate upwards but outwards too throughout the ceramic CPU package. The excess paste that gets squeezed out from between the CPU and heatsink on to the rest of the package then heats up and traps the heat there. If you look at the photos of the excess paste you will notice that the paste was bubbling away (boiling) from the heat. You can see that from the small pitting in the surface of the excess paste. That heat would have flowed away freely in the air if the excess paste wasn't there to catch it. That is why PC builders never liberally apply thermal paste.

[MacosNerd]

Originally Posted by indigoimac 

Ok, well this is interesting. Anyone crack open the latest revision 15"? I'm not about o but mines running at that same 59-60C idle in OSX, it'll kick up to close to 70-75 under load and maybe, just maybe the fans'll kick on. The Nvidia GPU hangs at about 10C hotter than the CPU.

My C2D iMac idles around 45C, I know it's a way larger enclosure, but it's still concerning that it's a 15C difference.

Umm the OP

59-60 seems to be the typical operating temps. I wouldn't be too concerned.

The 45C difference is largely do to the fact that the iMac is larger, can support more fans and those fans can run faster so heat dissipation will be greater.

[iampivot]

I'm wondering if the GPU can actually be the culprit in these new MBPs. I've got a week old 2.2Ghz SR MBP and my istat shows the GPU being about 8 degrees hotter than the CPU and about 6 degrees hotter than its heatsink when I'm doing mostly websurfing.



Maybe there's too much termal paste on my GPU?

[ghporter]

Originally Posted by PaperNotes 

Heat from a CPU doesn't just radiate upwards but outwards too throughout the ceramic CPU package. The excess paste that gets squeezed out from between the CPU and heatsink on to the rest of the package then heats up and traps the heat there. If you look at the photos of the excess paste you will notice that the paste was bubbling away (boiling) from the heat. You can see that from the small pitting in the surface of the excess paste. That heat would have flowed away freely in the air if the excess paste wasn't there to catch it. That is why PC builders never liberally apply thermal paste.

How can you "boil" thermal paste? What temperature are you thinking that this stuff boils at? Arctic Silver 5 is functional up to 180ºC. FUNCTIONAL. That means that it certainly doesn't BOIL at any temperature below that! What you're seeing in those pictures are air bubbles formed when the material was applied and squished out.

Further, THERE IS NO WAY A THERMALLY CONDUCTIVE MATERIAL CAN TRAP HEAT IN ANY WAY. That's like saying a screen door traps air.

[MacosNerd]

Originally Posted by ghporter 

Further, THERE IS NO WAY A THERMALLY CONDUCTIVE MATERIAL CAN TRAP HEAT IN ANY WAY. That's like saying a screen door traps air.

Well I do have an excessive amount of screen doors on my house and I did notice that my living room is getting warm now

One thing I am wondering about the reapplication of thermal paste is this; Apple (or who makes them for apple) in all likelihood does not use the top of the line stuff where but the OP and others who redid the thermal paste did, say using AS5. Those products probably have a superior heat transfer ability so the savings that we're seeing posted is a result of the product doing a better job and not because it didn't ooze out the side.

[tmoney468]

^you're right about AS5 having a superior heat transfer ability, but to say that gobbing on the thermal paste DOESN'T affect the temperature is just wrong. It's a fact that if there's too much thermal paste on the CPU, the CPU temperature is higher than if a small amount is applied. The main reason behind this is that this extra thermal paste serves as a barrier between the metal-to-metal contact on the CPU and heatsink. Thermal paste is only meant to fill the microscopic imperfections in the CPUs surface and the heatsink. In other words, if you want lower temps, put a small film of thermal paste on the CPU, as this is the ideal application.

[ghporter]

Originally Posted by tmoney468 

^you're right about AS5 having a superior heat transfer ability, but to say that gobbing on the thermal paste DOESN'T affect the temperature is just wrong. It's a fact that if there's too much thermal paste on the CPU, the CPU temperature is higher than if a small amount is applied. The main reason behind this is that this extra thermal paste serves as a barrier between the metal-to-metal contact on the CPU and heatsink. Thermal paste is only meant to fill the microscopic imperfections in the CPUs surface and the heatsink. In other words, if you want lower temps, put a small film of thermal paste on the CPU, as this is the ideal application.

I see evidence in the posted pictures of good contact between the heat pipes and the chips. There is no evidence that there was ever more than a very thin layer of paste between the chips and the heat pipe. Perhaps, as MacosNerd points out, the factory paste is far less efficient than the high-dollar aftermarket materials being used as replacement paste by those people who have disassembled their computers. Or maybe those people did not provide as effective an environment for their computers to cool in the ambient air, and after they went to all that trouble and effort, they provided (purely unintentionally) better airflow over the computer case and generated their lower temperature results that way. I would REALLY like to see a properly structured scientific study of this, with exactly replicable conditions both before and after. I'd also like to see a serious comparison of the factory thermal paste to aftermarket pastes in terms of thermal conductivity.

With the way the heat pipe is made, it is impossible to get a "horrible" fit between its copper pads and the CPUs and GPU. It won't fit properly within the case if it's not screwed in all the way, which WILL ensure that only the tiniest of spaces could possible exist between the chips and the heat pipe.

[nikhsub1]

Originally Posted by ghporter 

There is no evidence that there was ever more than a very thin layer of paste between the chips and the heat pipe.

Orly? Here is the SAME picture in my OP, but this one is actual size (cropped of course), tell me that is just a THIN layer on the chips?

http://www.anonforums.com/builds/mbp/largebefore.jpg

[MacosNerd]

Do you mind posting a smaller picture because that's really too big and messes up the thread

[stwain2003]

Originally Posted by MacosNerd 

Wow, I thought this thermal paste issue was a thing of the paste.

hahahahahahahahaha.

[olePigeon]

Are the MacBook Pro's CPUs socketable?

[bcaslis]

I agree with ghporter's posts. I have a B.S.E.E. and have dealt with lots of electronic components. I have no doubt that some of the people (the OP in this case) can do a better job than what the manufacturer assembly has done. However, I see no (as in zero, none, nada) evidence that the original thermal compound wasn't working. As long as the physical assembly squeezes out the excess there is no issue. Let's say that again, no issue. And if the physical assembly is not good enough to squeeze out the excess, then the OP's assembly would be even worse. if the gap was that large then the small amount used in the the photos might not be enough to make good contact.

I'm sure there are some machines assembled wrong and have over heating problems. But incorrect assembly is different than excess thermal paste. There is no evidence that the extra thermal paste is an issue except for internet experts that are frankly wrong. I'd much rather have a little extra paste than not enough.

[iampivot]

Any theory why there's a 6-7 degrees difference between the GPU diode and the GPU heatsink, as istat indicates?

[ghporter]

Originally Posted by nikhsub1 

Orly? Here is the SAME picture in my OP, but this one is actual size (cropped of course), tell me that is just a THIN layer on the chips?

http://www.anonforums.com/builds/mbp/largebefore.jpg

It appears that the original paste was as thin as torquing the heat pipe onto the logic board could make it.

Want to pay an extra $1000 for each MBP so that they're hand built (perhaps on the thighs of young virgins, like old-time cigars were supposed to be rolled on)? Unless the machine is dysfunctional, there's no reason to worry about the thermal paste. If it IS dysfunctional, TAKE IT TO APPLE AND HAVE THEM FIX IT. Otherwise why stress over it?

Originally Posted by iampivot 

Any theory why there's a 6-7 degrees difference between the GPU diode and the GPU heatsink, as istat indicates?

Higher quality paste and manually assembling the heat pipe onto the assembly could easily create that difference. Again, I have seen NO/ZERO/ABSENT/NONE evidence that "excessive" thermal paste is a problem. What I have seen is better performance from hand-assembled computers.

[nikhsub1]

Are you suggesting these machines are MACHINE assembled? ROFL! Not a chance. These machines are hand assembled like all OTHER PC's. And I'm not stressing over anything TYVM, I fixed my MBP's problem

[ghporter]

Originally Posted by nikhsub1 

Are you suggesting these machines are MACHINE assembled? ROFL! Not a chance. These machines are hand assembled like all OTHER PC's. And I'm not stressing over anything TYVM, I fixed my MBP's problem

There is a major difference between "assembly line" manufacturing and hand assembly. If my post did not make it clear that my reference was to the difference between high-volume, assembly line manufacture and the slow, patient, and painstaking assembly of a custom computer, then I apologize. However, you cannot be suggesting that there is any equivalence between the quality of an assembly line's assembly process and slow, patient hand assembly. Even with carefully torque limited tools, assembly lines live by FAST assembly, not careful assembly.

I for one believe that one part of this "issue" is the way the assembly manual is written. If it says "between .2 and .3cc of thermal paste", does it do so in such a manner that indicates whether this is supposed to be "per chip" as opposed to "total?" I bet not-and that could definitely make a vast difference between "OMG WTF BBQ!!! There's tons of paste in my computer!!!!!" and "There's a little paste squished out around the die on my computer's chips."

[nikhsub1]

I had a hunch, a theory if you will. I took my logic board out YET AGAIN... why? To test my theory. First, with my THIN layer of thermal paste, remember I used this special diamond paste which was VERY thick, almost like cold peanut butter. Well, the chips were making good contact with the sinks, no problem there at all. Now for my theory - I would bet anything that the thermal paste being used is Shin-Etsu G751, this is my paste of choice for my watercooled desktop machine. This paste DOES NOT AND WILL NOT SPEAD it is too thick. Intel spec for desktop mount pressure is ~50 to 70 lbs. With that amount of pressure excess will ooze out no matter the paste's viscosity. I dont know what Intel's mount pressure spec is for bare die CPU's (Desktop CPU's have an IHS - Integrated Heat Spreader covering the die, our notebook cpu's do not) but I can assure you it is no where near that of a desktop. If I had to guess I would say no more than 35 lbs. Anyway, with such little mount pressure excess thermal paste, especially the stuff Apple is using, is not going to easily ooze out. Again I believe this paste is Shin-Etsu G751 which many many OEM's use as it is very good and lasts a long time. So, if the stock paste is slopped inbetween the chips and the sinks some will ooze out, but not all of it! Please go back and look at my link to the large picture - note how much paste is on the chips! WAY too much! Here is a picture of the thermal paste on a desktop CPU after testing a bowed waterblock - the block was convex, you will note the strip going from top to bottom just off center - note how the paste was squished out! This paste in the below picture is Ceramique.



Anyway, tonight I removed the diamond paste which was also thick, not quite as thick as the stock stuff but still very viscous. I replaced it with Ceramique! I was fairly liberal with it (to test my theory) and my temps are now BETTER THAN EVER. Hard to give a number just yet as I just loaded the snot out of it with Intels Thermal Analysis Tool in windows.

To reiterate my theory once more - The amount of paste being used by Apple is NOT THE PROBLEM as people have stated, what is the problem is the viscosity of the paste, it is far too thick to ooze out from between chips and sink leaving a much thicker amount of paste which will hinder cooling in a bad way.

That's my story and I'm sticking to it

[chipchen]

Hmm... no one else thought it was a problem that he (the original poster) laid his BARE LOGIC BOARD ON CLOTH?

You might be worried about thermal paste... and thus shortening the life of the CPU... but what about the INSTANT DEATH OF THE CPU DUE TO ESD? You just don't treat circuit boards that way.... ever. It's like rubbing RAM on your pants before installing them. Hopefully you're not going to kill your MBP, then bring it to Apple for a repair.

Here... googled "ESD safety" and found this: http://resource.intel.com/telecom/su...ig/esd/esd.pdf

You (the original poster) should read it. Also the