Thin tape over gpu

I am currently repasting my GPU, before I put this back together I just want to make sure this is normal. On the GPU there is a thin film around the processor, but under it there looks like there is quite a bit of gunk, is this something I should pull off and clean or just leave it there?

Thanks for your responses.

 

Not my picture its from the forum im pretty sure you are talking about this?


Photocred* Wasaki

 

That is correct, I have repasted a few times now but only achieved around 60c idle - 95c load, I am kind of curious if that stuff might be the reason I can't get better temps. (20-23c ambient)

 

You can remove this without any problem, but it won't make any difference regarding temps.
Some manufacturers have this around the die, it's mainly there that the paste doesn't get everywhere.

 

Make sure you are using IC7 Diamond to repaste with it sets like concrete and I have tried every paste under the sun and never got below 90 until I use ICD.

I posted my latest results here it just runs gobsmackingly cold 82oC max under extreme burn with all settings maxed and my clocks below.

Dont expect much difference on the CPU.

 

That's actually fairly normal - I've got a thread in my sig with links to about 15 other people who have repasted and still end up with 95s. Myself included.

So there are a few people who get amazing temps, but that is *not* the norm.

 

What paste and method did you use if you dont mind me asking? Or are you in a hot tub in the Bahamas with 3 russian ballet dancers, always a possibility

 

Alright, I actually noticed something this time, when I was screwing the GPU heatsink screws in, #1 could still be moved up and down even thought it was as tight as I could screw it in (without applying a ton of pressure), while the others wouldn't budge when screwed in.

 

Try shifting the heatsink round a bit and make sure nothing is blocking the hole the screw goes into.

If not that means it is not threading the actual screw should not come out of the socket its spring loaded. Are you able to move the heatsink at the corner where screw 1 is? If so that is bad news. However I believe that 1 is the top left hand corner?? and I found the heatsink tilted upwards at the bottom so you may be ok with the other 3 screwed down however its not good recommend you get a replacement crew. If it is not air tight bubbles can appear at the corner and it was what happened with mine because ASUS did not even bother screwing one corner in properly and it burned upto 102oC.

 

I am too amused that Dallers is a retelling of my own story

 

Probably the same bloody tech that pasted both of our laptops thats why

 

You may remove thin foil and put back again just to clean under but its best to leave alone if it is not to nasty. And foil is there for reason, its static shield to protect capacitors sitting on gpu chip from sparkling with copper heatsink.

 

That's your opinion.
And try to run Furmark with either higher resolution or better fullsreen in native resolution.
No CPU load tools simultaneously.

A MSI 5870m don't needs to be hidden.

 

Thats not my opinion its what the results by others show and the resolution means nothing when Im testing different types of paste with the same settings. Also the CPU is totally separate to the GPU cooling system and as such would have minimal effect anyway but that doesn't matter because again using the same testing method for each paste would produce no difference in the results.

I used exactly the same testing method with Arctic Silver and MX3 5 separate occasions with 5 separate ways of pasting and it did not drop under 90oC with the results posted on Chastity's repasting page.

I was ignorant to IC7 at first and have used Arctic Silver 5 for many years ignoring what others said about the IC paste and I only applied it once willy nilly and the results I got destroyed any other paste I have used on any repaste I have conducted.

Try it and see for yourself.

 

I don't want to blame you. But the fact, IC is the only supa dupa tim, isn't right. All are using this stuff because one recommend it.
Who and where are the others??? I'm the opposite of the others.
Maybe they have done some mistakes?

I know how this cool system is build.
The CPU thing might be, that furmark won't get enough cpu power while other tools stresses the cpu to 100%. Running Furmark alone uses 1 core at 100%!

OK, i've redone a test with my MX-3 pasted buddy at your settings.
To compare with!

700/1000


825/1100

At 83°C the fan kicks into turbo state (4200rpm) and comes down to 3700 again.

70/100 Idle (fan stop at 43°C)

 

My recommendation is different to yours that is all and not once did I say it is the only super duppa TIM only that my results are that of testing all three. So how can you claim im not right lol when I didnt even make that claim. Also would love to see how you came to the conclusion that only one recommended IC Diamond when there are literally thousands of people saying exactly what I am.

8 minutes at my clocks in your test that barely reached peak and 15 minutes on stock clocks? how is that a fair test??. 15 mins + and then repost again. Also my idle clocks where at 405/1100 not underclocked as yours are.

 

I claim, that you must have done something wrong while testing with the other tims. Thats all. You can't get 10+ lower temps with just replacing the tim.
Another 15+ minutes won't change this. But yeah, you can stress your GPU all day long.

 

No becouse static foil sitting on heatsink of msi if i remember correctly

I have to agree with you tilla. I never seen more then 3-5 degrees in difference no mather what kind thermal paste has been used on same machine.

 

OK. That's right. I missed that part.

 

Lol, np. Dont get sad now .

 

Another thread degenerated into pointless TIM argument.

 

+1. Correct answer

 

Thanks.
It was still somewhat interesting.