How I Lowered My temps for £3.25

http://img811.imageshack.us/img811/7...foreafterg.png

Good Thermal paste Pays Big time. the Chinese stuff would be better sold as house insulation

Same load to heat it up?

Was the stuff spread a cm thick or something!!?

kinda pointless unless both pics were at load or no load, they are not the same ;)

both were at idle . the jump was from a momentary app load. both were applied in the exact same way. rice sized drop on the heat sink base, half rice sized drop on the ihs. the first capture was taken almost immediately after boot . the second was after about 3 minutes. i could not get a temp at load due to the computer shutting down almost immediately with the cheap paste.

looking at core 1 the heatsink may not have been on correctly m8

Either way the lowest temp on the chip was 68, which tbh was pretty high. Compared with the highest temp of 40 means that the 3 minutes time difference doesn't really matter.

Toothpaste and vegimite are meant to work the best. A bit of water in there works wonders apparently.

I have to agree with Gonzo; I seriously doubt you can get a 40C difference by changing paste. What was the GPU usage during the two measurements?

What was the 'bad thermal paste'? OEM with hsf?

There's no way the heatsink was installed properly first time round. Going from the worst OEM thermal paste to the best pastes can't make a 10c difference, let alone 50c.

the higher temp in core 1 i think was due to the heat sink being a little loose in that corner. the thermal paste came out in clumps and because my heat sink is direct contact could not fill all of the nooks and crannies around the heat pipes. i use it when i need to make sure a cpu works it's never on for long and it really is very bad

A bit of Arctic Silver 5 is always my go to. I still have some kicking around - I'm about to buy an Arctic Freezer 7; reckon I should use the already applied stock paste or clean it and replace it with Arctic Silver 5? (Since it's an arctic cooler, I wonder whether it's a decent application of Silver 5 anyway, in which case my poor thermal paste application skills will definitely rear their ugly heads and ruin everything).

Basically - go with the perfectly applied stock paste (which could well be arctic silver 5..) or inevitably badly applied arctic silver 5? Hmm...

EDIT: Nevermind, answered my own question - http://forums.anandtech.com/showthread.php?t=144592

pre applied is liable on putting air bubbles between, unlike a blob in the middle of the AS5 that will spread out with no air giving better results. also the spreader on chips is never flat making the pre applied pointless apart from the lazy who want to make do :)

i think the preapplied paste on arctic cooling hsf's is MX2 which is a pretty damn good paste as is i would stick with that

Yup, definitely a badly seated heatsink :) Congrats on your lower temps. Next step: undervolting! Cost: £0.

actually I once had similar results.

And I did it three times and it was fitted right.... it was just rubbish paste that clearly insulated the temp... didn't conduct it.

In my cae it was a white pasty stuff that looked like poly filler.

it was rubbish and my pc only held Windows for a few minutes before crashing.