Temporary effects of using no thermal interface material?

Xylo

Limp Gawd
Joined
Jul 5, 2005
Messages
421
Very random question, but, what effects would there be, if any, of not using a thermal interface material like ArcticSIlver or similar for a few days on a CPU/HSF setup?

I have a clean HSF and a clean CPU, and I'm going to eventually be watercooling the CPU, but before I do, I want to double check a few things first before I take the time to mount everything...however, I really don't feel like wasting more ArcticSilver and then cleaning it all off for what might be 1-3 days of use...

I can't imagine it could be unsafe to use no thermal grease temporarily, but figured I'd double check anyway... :)
 
DermicSavage said:
I wouldnt run without thermal grease for more than about 3 seconds

Really?...why?...Does it make that much of a difference in performance temperatures?...I didn't think that it did (especially before given time to cure)?...
 
i wouldnt take the chance on it, could cause some serious damage, the thermal paste is what creates a bonded contact between the HSF and CPU, and although its on pretty tight, the HSF, without it you could do some serious damage, those are my 2 cents.
 
on the first computer i built, i had a 1900+ and a HSF mounted on with NO thermal stuff because i didnt know you needed it. It ran perfect until about 15 minutes of playing tribes 2, then it just hard froze. It took 3 more freezes of that type to finally make me understand that i needed thermal grease. i applied some and the computer has been running fine since then. 4 or 5 years? The only damage that happened was that some of the text on the processor core was imprinted onto the heatsink, that was it

i dont suggest you use no thermal paste though
 
Xylo said:
Very random question, but, what effects would there be, if any, of not using a thermal interface material like ArcticSIlver or similar for a few days on a CPU/HSF setup?

I have a clean HSF and a clean CPU, and I'm going to eventually be watercooling the CPU, but before I do, I want to double check a few things first before I take the time to mount everything...however, I really don't feel like wasting more ArcticSilver and then cleaning it all off for what might be 1-3 days of use...

I can't imagine it could be unsafe to use no thermal grease temporarily, but figured I'd double check anyway... :)

what type of proc are we talking about here? I've been running an athlon 1.0 for about 5 years without it.

places like monarch make thermal grease an option when purchasing a 939... so it seems that it is ok.

let me end by saying this before i get fully scorched by the flames... i wouldn't reccomend running a modern PC without it.
 
As in all things I suppose it depends on a variety of factors, but why take the chance? IF you had a very finely lapped HSF unit (down to say 2500 grit or so) and your CPU was also super smooth it *might* work. Please keep in mind that thermal interface material is designed to fill up the microscopic pores of the metal in the HSF so you get a much higher degree of actual surface to radiate heat from. If you go to the Artic Silver website they have some nice write-ups and pics about this.
 
it's possible to do so with the newer proc's, they have heat protection ;) but it won't run long tho'

but with the older cpu's it aint advicable, they would just die under your hands ;)
 
I made 2 pic of reason y u should apply some thermal paste to the cpu.




any question?

Note: Hotspot is where the heat stay and possible expand the silicon material.
 
BigDaddy85 said:
I made 2 pic of reason y u should apply some thermal paste to the cpu.

Heehe...thanks for the pics. :)

1) I understand the point of using thermal grease in general, just wasn't sure if it mattered short term...I guess it does/

2) What's "craved"?

3) I guess I'll just be using some, then cleaning it off yet again...I can't help but to wonder though if I'm really scraping it up bad at a low level with the cleaning (will be it's 3rd time now, I think?).
 
If it were me, I'd probably do it just to prove everybody wrong. Of course you need it, but I bet you could get by without it. I'd also underclock/undervolt the processor right off the bat.

Of course I guess it all depends on the processor. I'm instinctively thinking A64 here.

If you want to take the middle road, use the arctic silver you have but don't use a lot. Just scrape a very thin layer out of the tiniest dab and it'll make a difference. Then you won't waste much.

Option 3: Are you sure you don't have any tubes/packets of generic thermal compound laying around, from other purchases? (heatsinks, retail cpu's, etc) I have a bunch.
 
Pft, I'm going to do this when I get my motherboard. I'll run some benchmarks and record temperatures with no thermal interface compound, then I'll try it with AS3 and see what the difference is.

:rolleyes: @ the people who don't even know what the difference will be but assume it's going to be dangerous
 
In the small tube of AS-5 (3.5 grams) you have enough material to build about 30 or so computers. We are talking about one tiny drop smaller then a small “BB”, so I’d think you can probably “afford” to waste a bit as cheap insurance. ;)
 
thread Hijack/
I got a Gigabyte cooler and it came with Shin Etsu grease and I squirted the whole thing on the proc and installed the cooler......anything gunna go wrong?


/thread Hijack
 
i have always ran with it... the processor chip and the heatsink will never ever match up perfectly flat and the paste fills the void... and feel free to proove us all wrong if you got the cash to burn (or to not burn... you will see).... more power to ya!
 
im running my processor right now with out thermal grease? the tube that came with my heatsink got smashed and all the grease came out, ive been running it like this for I dont even know how long.
 
Being a mechanical engineer i would say that thermal grease is paramount when cooling a CPU. However, it is not necessary if you have a perfect interface between sink and cpu because the thermal contact resistance would be infinitly small. THis is not the case when dealing with cast and then machined heat sinks. The sinks have some thermal contact resistance due to the roughness of the interface itself. THe thermal grease lowers the contact resistance between the sink and cpu by filling the gap. As we all know air has low electrical conductivity hence low thermal conductivity and acts as the major bottleneck of heat transfer
Hope this helps.
rover
 
Xylo said:
What's "craved"?
Just like wat rover mentioned recently. Metal surface isn't perfect flatted out so there will be several "gaps" on it. If u have microscope then zoom in and u will kno wat rover's talkin bout. ;)

This is the reason y the thermal grease or pad exist today. :p
 
Being a mechanical engineer i would say that thermal grease is paramount when cooling a CPU. However, it is not necessary if you have a perfect interface between sink and cpu because the thermal contact resistance would be infinitly small. THis is not the case when dealing with cast and then machined heat sinks. The sinks have some thermal contact resistance due to the roughness of the interface itself. THe thermal grease lowers the contact resistance between the sink and cpu by filling the gap. As we all know air has low electrical conductivity hence low thermal conductivity and acts as the major bottleneck of heat transfer Hope this helps. rover

He's absolutely correct.

The best contact is metal to metal if all over variables being correct, its a 100% contact interface from material to material.

If you can insure that the surfaces are absolutely 100% mating, then nothing can beat metal to metal contact.

The purpose for pastes is to fill the voids in machining, and manufacturing flaws. Its a crutch for the fact that machining and materials have inherint limitations in their abilities.

I have my P4 1.6 @ 2.4 running just fine 24/7/365 (it never gets shut off, and it doesn't crash) and *Gasp* I didn't use grease. How is that possible you say? It just is :)

I agree though, paste is mandatory, but thats not to say because you fail to use it hell is going to burn around you. But I do agree, AS and those types of things serve the purpose of fixing poor machining and filling the gaps.
 
stiltner said:
I agree though, paste is mandatory, but thats not to say because you fail to use it hell is going to burn around you.

Exactly right. It's really frustrating how many people will reiterate ye old "microscopic valleys" speech before they realize they aren't really adding anything new to the discussion. It's a sidepoint, not an answer to the to the question, and gives a false impression when presented as such.
 
if you buy a cpu that comes with a heatsink and fan will the thermal paste already be on it?
 
r3m3dy said:
if you buy a cpu that comes with a heatsink and fan will the thermal paste already be on it?

it will have a thermal pad, just remember to remove the plastic cover
 
Wildace said:
it will have a thermal pad, just remember to remove the plastic cover

I think I forgot to do that on my last install!!!

The prescott P4 & Celeron D (retail) stock HSF's have a greyish TIM with a plastic cover right?

I noticed the temps were kinda high, but didn't think about it until much later. I need to go make a house call on that guy.... bring some thermal paste and check. B/c he's probably got a melted plastic thingy on the HSF!!! DOH!!!!
 
BillR said:
In the small tube of AS-5 (3.5 grams) you have enough material to build about 30 or so computers. We are talking about one tiny drop smaller then a small “BB”, so I’d think you can probably “afford” to waste a bit as cheap insurance. ;)
QFT..... the amount of paste needed to protect the average CPU is so little, it would be ludicrous to run without it, JMHO.....
 
coolie_d said:
QFT..... the amount of paste needed to protect the average CPU is so little, it would be ludicrous to run without it, JMHO.....

Well, no. It's just a matter of cleaning it. I hate having to clean the stuff off, then remount, etc...like I said, I wonder how much scraping of the surface I'm doing cleaning the stuff off... :/ ...I *try* to be gentle....
 
Xylo said:
Well, no. It's just a matter of cleaning it. I hate having to clean the stuff off, then remount, etc...like I said, I wonder how much scraping of the surface I'm doing cleaning the stuff off... :/ ...I *try* to be gentle....
with a nice soft lint-free cloth and some 90% isopropyl, I find it is really easy to clean the stuff off (works awesome on the A64s and P4s because of the heatspreader), just make sure you use 90% and not the lower concentration "rubbing alcohol"..... always works great for me....

Edit: if you are in fact using thermal paste, and not thermal epoxy or something similar, there should be no scraping involved at all....
 
Dear LORD !!! For love of God anyone NOT willing to throw their processor in the garbage can and is reading this.. Put the thermal interface material on !

"air has low electrical conductivity hence low thermal conductivity"

In the context of this discussion this is basically correct. However;

Air is not highly conductive but compared to many things (rubber for example) it is actually a decent conductor. My major issue is the inference that low electrical conductivity infers low thermal conductivity. Pure water comes to mind immedaltely, a very poor electrical conductor but a superb conductor of heat. Pure water was and still is, I think, still used to cool extreamly high powered transmitter tubes. Note that "pure" in this case means laboratory pure.
 
Back
Top