![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
-
Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.
Thermal compound optional?
- Thread starter aironite
- Start date
- Status
S
SlickJesus
Guest
no thermal compound = burned up CPU
BlindedByScience
More Human than Human
- Joined
- May 26, 2000
- Messages
- 9,282
Your local Radio Shack has "Heat Sink Compound" that will work in a pinch. Your temps will be within a degree or two of the "designer" thermal compounds out there.
Assembling your CPU / HSF without a thermal interface of some type is bad news. Don't do it......
B.B.S.
Assembling your CPU / HSF without a thermal interface of some type is bad news. Don't do it......
B.B.S.
P Masta Flex
Weaksauce
- Joined
- Sep 27, 2003
- Messages
- 77
artic silver 5 is great.
id like to see what kind of difference it actually makes..
im willing to bet not as big as you guys are making it out to be...
i say try it... no way in hell your proc is going to burn up...
unless your h/s is concave or something.. in which it would get horride temps anyway..
the thing will lay flat on the processor...
im willing to bet not as big as you guys are making it out to be...
i say try it... no way in hell your proc is going to burn up...
unless your h/s is concave or something.. in which it would get horride temps anyway..
the thing will lay flat on the processor...
the_village_treky
Gawd
- Joined
- Feb 3, 2004
- Messages
- 1,009
thermal compund isnt realy THAT importatn. a good car analogy would be oil or gas
lol, yes you need it, and get the good stuff, like as5 or as:cAnd how will heat transfer, then, eh? Shim to HS?
Ok, so lets assume it does. Now then, heat transfer is proportional to surface area for most normal substances. That means, the more surface in contact, the better the transfer. Air is an insulator at this level of transfer (within a solid/liquid substance). Now then, lets look at this from a molecular level:
There will be a grand sum total of two to three molecules here in contact. Neither surface is perfectly smooth, so there are two to three molecules in contact from each side, the rest separated by a miniscule amount of air (which, as we decided earlier, is an insulator). Now then, the cross section of a molecule is, well, rather small. For the sake of argument, we're going to assume that it's, oh, say .05cm, which is HUGE in comparison to a real molecule. Now, the base of your HS is, oh, say 3cm sq. With thermal compound filling in the entirety of the gaps in our previous situation, you get the full 3cm sq to transfer heat. Almost as if it was one object. In YOUR idea, you get .0025cm sq to transfer heat. Now, going back to our proportionality, this means that HSF with compound transfers heat 1200 times to 400 times better than without. And that's assuming an absurdly large sized molecule.
Now, even if they are close enough that more than one point comes in contact, you're still dealing with a miniscule amount of transfer.
You need the compound.
EDITED for an error correction
jamestime88
[H]ard|Gawd
- Joined
- Dec 10, 2003
- Messages
- 1,209
not the BEST advice i have ever heard
a compound increases the efficiency.. i do not argue that point..
but by what %?
you guys make it sound like its over 80%.. which is way off.
there is plenty of surface area on the hs to lay on contact with the heatspreader or die..
im not saying this would be the best case scenario however... im saying that it would work. w/o a processor burning up or a system shutting down.
but by what %?
you guys make it sound like its over 80%.. which is way off.
there is plenty of surface area on the hs to lay on contact with the heatspreader or die..
im not saying this would be the best case scenario however... im saying that it would work. w/o a processor burning up or a system shutting down.
BlindedByScience
More Human than Human
- Joined
- May 26, 2000
- Messages
- 9,282
....and if this was really true, do you think for a second that Intel / AMD wouldn't stop including TIM's of some flavor and save 0.5 of $00.01 on every CPU they sold (retail).....??
To say that "...there is plenty of surface area..." for the thermal transfer forgets the fact that when the CPU designers did the thermal model for the package and die, they included a thermal transfer material between the CPU's heat spreader and the HSF when they did the calculations on how the die / package / heat sink / fan / surrounding environment would work. Leave it out = things get WAY too hot. Also, consider that at the microscopic level, due to surface roughness, there's not as much metal to metal contact as you may initially think. Keep in mind how good a thermal insulator air is when it's trapped and can't move....like in every little microscopic void between the two.
They don't throw TIM's in there out of the goodness of their heart, they are there so they don't have to warranty CPU's. But of course, don't take my word for it - by all means, give it a whirl.....
Mad_Pyro: I got two degrees C going from Rat Shack to AS Ceramique. My temps still just kick ass but my numbers were based on personal experience. I'm willing to bet that if you had a hotter CPU, you could see a bigger difference (I didn't totally doze off in thermo....
) Another thread......Regards - B.B.S.
Buddy, I showed you mathematically. It's 300-1200 times, assuming an absurdly huge molecule cross-section. Now, in reality, it's several ORDERS OF MAGNITUDE more.
Either read the posts and respond, or don't respond at all.
How much of that HS do you think will actually touch the CPU? You think it will lay flat? Try again... Try THREE molecules. Or groups...
Lets put it mathmatically again: You bring the two together. The first point hits (molecule or two wide, since the HSF and cpu are VERY rough at the molecular level). The HSF tilts microscopically till the second one hits. Now you've defined a line of contact, by two points. When the third point hits, you've now defined a plane of contact, by THREE points. THREE molecules (or small groups). because it is an uneven surface, that's all that touches. If there were any more out past those three points, they would have already touched. All remaining points on the plane are farther away, and thus do NOT touch. Air is an insulator, it transfers very little heat when you're talking about something that runs HOTTER THAN A NUCLEAR REACTOR (in terms of watts/unit area) especially in comparison to TC. So, you've got 3 points trying to transfer heat. The result: It doesn't work well. You're over a thousand times less efficient.
They aren't perfectly flat. They don't lie together perfectly. That's what TC is for. It fills in the gaps, removes the insulating layer.
Originally posted by BlindedByScience
....and if this was really true, do you think for a second that Intel / AMD wouldn't stop including TIM's of some flavor and save 0.5 of $00.01 on every CPU they sold (retail).....??
To say that "...there is plenty of surface area..." for the thermal transfer forgets the fact that when the CPU designers did the thermal model for the package and die, they included a thermal transfer material between the CPU's heat spreader and the HSF when they did the calculations on how the die / package / heat sink / fan / surrounding environment would work. Leave it out = things get WAY too hot. Also, consider that at the microscopic level, due to surface roughness, there's not as much metal to metal contact as you may initially think. Keep in mind how good a thermal insulator air is when it's trapped and can't move....like in every little microscopic void between the two.
They don't throw TIM's in there out of the goodness of their heart, they are there so they don't have to warranty CPU's. But of course, don't take my word for it - by all means, give it a whirl.....
Mad_Pyro: I got two degrees C going from Rat Shack to AS Ceramique. My temps still just kick ass but my numbers were based on personal experience. I'm willing to bet that if you had a hotter CPU, you could see a bigger difference (I didn't totally doze off in thermo....
Regards - B.B.S.
That's what I said
!
BlindedByScience
More Human than Human
- Joined
- May 26, 2000
- Messages
- 9,282
Wow, another speed reading miss....me = pwned...Originally posted by lopoetve
That's what I said
< end /threadcrap >
Regards - B.B.S.
I do agree that TIM's do no doubt help with temps, and it's always a good idea to use them, but by how much do you think they really help? My guess is it wouldn't be enough to get up to the CPU throttling levels.
This is just a guess, so don't take my word for it or anything, but I'd bet if someone booted up without TIM it would probably still within operatable temperature levels (at stock speeds anyway).
Yes, in theory, it's something you can't live without. In reality? People have done without it before...
This is just a guess, so don't take my word for it or anything, but I'd bet if someone booted up without TIM it would probably still within operatable temperature levels (at stock speeds anyway).
Yes, in theory, it's something you can't live without. In reality? People have done without it before...
S
SlickJesus
Guest
I'll take you back to a time when i owned a pentium 3. i put my hsf on without using thermal compound. do you know what happened? thats right, a burned up CPU.
Lesson here? if you don't use TC on a P4 you will reach clock throttling temps, do it on an athlon and you have a dead athlon.
S
SlickJesus
Guest
celerons definately don't put out as much heat as P4s
you used to be able to get away with it, but i wouldn't even try it anymore.
you used to be able to get away with it, but i wouldn't even try it anymore.
Originally posted by RS3RS
Ah, well, you could probably tell I wasn't exactly sure
I guess I was just going on what I remember from running my old celeron 333 without TIM for months, but I doubt those things put out as much heat as a P4...
Technically, a P4 puts out more heat per unit area than a nuclear reactor does.
I used to run old Celerons/a P233 without TC. Doesn't work now.
Now now, no need to get saucy. It just happens that your opinion is wrong. That's a matter of fact and science. Don't get so upset when your wrong it happens to everyone a lot in life.
On a lighter note I love posting this article in Thermal discussions.
http://www.hardocp.com/article.html?art=MTQ0
It's getting a little dated, maybe time for a refresher... But it does make good points. BTW when comparing to 'current' designer goop just consider the manuf. chart's of imporvements vs. their prior designs, it's usually minimal from version to version....
OK, anyone who disagrees with the FACT that thermal compound is needed do this. Remove HSF from CPU, clean both and dry. Reinstall CPU and HSF sans thermal goop. Boot machine and run Prime95 torture. Have fun replacing equipment or testing the thermal protection in the CPU.
It's like arguing that a brick isn't solid, or water at room temp. isn't liquid. There is science behind this, no opinions as Lopoetve said. They only real discussion is the effectivness of various products ranging from TIMs, to white goop, to expensive designer compounds.
It's like arguing that a brick isn't solid, or water at room temp. isn't liquid. There is science behind this, no opinions as Lopoetve said. They only real discussion is the effectivness of various products ranging from TIMs, to white goop, to expensive designer compounds.
kleptophobiac
Supreme [H]ardness
- Joined
- Sep 24, 2001
- Messages
- 7,839
I wouldn't put much credit to your mathematical proof, with more than a little fudge to your numbers.
The point is still valid, even though you pulled the data out of your ass.
Don't run without thermal compound unless an engineer at IBM used his scanning tunneling microscope to make sure your HSF and CPU were perfectly flat.
BlindedByScience
More Human than Human
- Joined
- May 26, 2000
- Messages
- 9,282
....question asked, opinions offered (...!!!!...), and getting a little off topic / personal....
B.B.S.
B.B.S.
- Status