Arctic Silver 5 temperature reduction poll

For those who replaced stock paste with AS5, without making any other alterations how significant was your temperature decline (average CPU and GPU decline).

 

This thread is quite timely. I was going to stop by the store tomorrow to pick up some AS5 for my XPS 1645. Are there any other alterations worth doing besides just AS5 alone?

 

Thoroughly cleaning the fan/heatsink.

Here's a picture of how nasty it can be:

If you're having serious temperature issues you could also do an undervolt of the CPU.

I replaced my thermal paste today it seems like there's a few degrees worth of reduction, but it takes a while for the paste to settle to experience the full effect. I definitely got more benefit from cleaning than the paste SO FAR.

It was my first time switching thermal paste, so I'm concerned I didn't put the right amount. I put a very thin layer on the heatsink and a similarly thin layer on the GPU/CPU and spread with credit card. (It was my second time doing it, yesterday I used the putting a pea sized amount and squishing the heatsink on top. It didn't spread out very well so I redid it. I'm still not certain about it.)

 

DAMN!!!!!! How many months was that buildup from?

 

That actually is not my laptop, but I've had my laptop for about a year and I noticed it was getting really hot and while I could hear the fan I couldn't feel the air coming out the back. After cleaning it I feel more airflow, and it's noticeably cooler.

I did end up getting a few pretty big dust bunnies out of it.

 

I got ambitious and pulled the heat sink off. I love how it's nice and simple. I can't put it back together until I pick up some AS5 in the morning. My heat sink is clean, the laptop is fairly new.

 

Make sure you only apply the paste to the actual chip and don't put any on the heatsink. If you put paste on both the chip and the heatsink you risk air bubbles forming and you won't have the best possible contact.

 

Arctic Silver recommends spreading AS5 onto mobile Core i7 processors, rather than the "line method". I thought this was ill advised as it can promote air bubbles. Link to Intel processor page.

 

Yeah, I realize now I misread the guide. The guide talked about tinting the heatsink where you apply the paste but then wipe it off to leave only trace amounts in the tiny cracks/imperfections. I did not read that part *sigh* and probably had half a grain of rice's worth on the heatsink and another grain of rice on the CPU/GPU.

If my temps are bad I will redo it, sigh. Do you think it's worth redoing it, or better to wait for the paste to settle and see what kind of temps I'll have?

Thanks

 

Don't even bother with AS5! Get ICD7 instead. It's non conductive and superior in every way. Reduced the temps in my laptop by 11C. Yes, 11C. It may cost a bit more but the difference is day and night.

 

I briefly read up on this product and it does sound very nice! I'll see if Micro Center carries it.

 

The day after applying AS5 (with the wrong tinting the heatsink method)

Idle temperature MIN/MAX (from HW monitor):

CPU: 41/50
Core0: 37/44 Core1: 33/42
GPU: 57/59
HD: 28/30
Fan speed: 0/0

(This is about 20 minutes after turning computer on, and it's about 62 degrees in my room. Laptop on flat surface.)

Load temperature (this is running orthos and rthdribl for 3 minutes)

CPU: 70/73
Core0: 58/62 Core1: 57/61
GPU: 71/74
HD: 31/31
Fan speed: 3680/4280


My video card is the geforce 8600m gt, and the cpu is the duo [email protected]

 

After a really long time of internet browsing (with up to 15 tabs open):

Here are MIN/MAX temps: (It's slightly elevated on a laptop stand, but its external fan was not on.)

CPU: 38/58
Core0: 33/52
Core1: 30/48
GPU: 54/65
HD: 33/38
Fan speed: 0/3240

It pretty much cycles between the min and max (reaches the max, fan turns on, reaches the min, fan turns off).




(starting at the highest point of the idle cycle):

8 minute orthos rthdribl load tests:

CPU: 80
Fan: 4600 rpm
Core0: 66
Core1: 66
GPU 77
HD: 39


(starting at the lowest point of the idle cycle):

3 minutes:

CPU: 74
Fan: 4280
Core0: 65
Core1: 64
GPU: 74
HD: 38

8 minutes:

CPU: 76
Fan: 4600
Core0: 67
Core1: 67
GPU: 77
HD: 38

 

I JUST applied the AS5 on my XPS 1647 about an hour and a half ago, i understand that it can take several heat up/cool down cycles and upwards of 200 hours (so i've heard) to reach it's break in period. but as of this, my temps for under load have not changed any for my GPU(78C on average) my highest i've seen before the change was 81C but i've not had time to let it get that high yet. but my idle temp for my GPU has increased by 5 or 6 degrees... i'll give it some more time to give a final verdict...

 

I hate to say it, but the break-in thing is going to be like a 1C difference.

But for the idle temperature, remember that it really doesn't matter. You should compare how often the fan came/comes on for measuring AS5's "performance" for idle. The reason for this is that the fan comes on and off based on the TZ01 temperature-- Suppose every laptop comes with Dell's stock paste on the CPU and GPU. Now imagine you put the world's best paste on the CPU and Dell's stock paste on the GPU- your GPU average idle temp will go up even though it's the the same stuff, because the fan will be on less frequently.

My point is it may seem worse, but it might not be. Perhaps your CPU temps are much better now, and for whatever reason the GPU is approximately the same.