Is there such thing as TOO MUCH?

I have read threads on what kind of thermal paste to use and high temps, first thing to look at is replacing the paste. I have also read different views on the ammount of paste to aply. Some say put a small spot on the die, others say to cover the dies completely with an even layer.

My question is, is there such thing as to much thermal paste? Obviously to little could lead to high temps. Could to much lead to high temps? Could to much lead to lack of OC ability? Thoughts anyone?

 

yes too much can actually be worse than having too little or even no thermal paste, and depending on its composition if it is conductive and oozes has been known to kill IC's

 

I got heavily into water cooling a couple of years ago, 5 water blocks on that rig! Like you I found a LOT of advice as to what paste, amounts, and how to apply. Two most common were five small dots in a dice pattern, and applying to one side and spreading it with a business card to coat the surface in a thin coat - which I went for. I also checked spread patterns by opening up again and looking, a barely visible coat at this point seemed to be the best. Different pastes only made 1 or 2 C difference for me.

 

Yes. Too much thermal paste is just as bad, if not worse than too little.

Picture it like this, your die and heatsink look smooth and probably feel smooth, right? Wrong - if you zoomed in far enough you would see both surfaces are rough, and this is bad when you are trying to conduct heat, as direct metal on metal contact is the most efficent way. This is why we put a THIN layer of thermal paste on, to fill the gaps and make the most efficent transfer area possible. If you put too much thermal paste on, yes you will fill the gaps - but now the thermal paste almost acts like an insulator, trapping the heat inside.

I wouldnt fuss too much about what thermal paste you use, nor how you get it on there, but I cannot stress enough you need to put the RIGHT amount of thermal paste on. About the size of a pea for a GPU and an uncooked grain of rice for a CPU is the right size and amount.

 

I can't go over 64C on my GPU, it's immposibe. So to much or to little isn't my case in temps right now. I appreciate the thoughts gentleman, but the question still stands. Can to much thermal paste be a cause in a lousy ability to OC your card?

 

Ahh, now I see where you are coming from

As long as temps are as low as you are seeing thermal paste is not part of the equation. You could just have one of those cards that is borderline or just does not play nice. I had an R3/580m that I could OC to 750 quite happily on stock voltages. This 675 (exactly the same card) breaks up above 720 .

Exactly what happens when you OC?

 

No, not really... its generally to do with your voltages. I had a card that overclocked bad, and I had to raise the voltage. Now it overclocks like a boss. 64 degrees is nothing and that shouldn't affect clocks.

 

Yeah, I have read all about simply getting a bad card for OCing. I'm just having a hard time excepting that, for how much I spent on it anyway. It's in my head I should be able to get a lot more out of it. I also realise that anything over stock clocks is a bonus. I guess I'm just reaching for any excuse other than "it's just a bad OC card"

When i OC "to far" I crash 3Dmark11 and get a bubble in the lower right that says "Driver has crashed and recovered" It's all the coming cases of the clocks not being stable. OV doesn't allow me to get that much farther, but still very happy with my purchase. I think I just expected more for the tag that is on the side of it.

 

64C is cool, don't think you should have any issues overclocking it.

 

I understand that, one would think anyway. But even OV'd and OC'd I couldn't get a temp over 67C. I crash on games or benching with to high clocks but nothing that extreme like I have seen from others.

 

That error (TDR) is quite common and can be just the driver in some circumstances. Have you tried an earlier driver? It's a windows timeout thing when checking the card is responding, and happens with ATI cards as well (the problem seems to be in the windows HAL).

 

Oh really? Thanks, I will have to look into this. One would assume that a newer driver would be "more stable" for such activities. But we all know what happens when we assume.

 

Yeah and I assumed I'd just fa****,

Latest NV WHQL is getting panned for this and other errors

https://forums.geforce.com/default/...ay-driver-feedback-thread-released-5-23-13-/1

 

Wow, nice link. I haven't tried a high OC on this latest driver but I will look into an earlier driver and maybe try again tonight.

BTW, did this WHQL driver fix your driver proplems you were having?

 

nah. I found that turning off hardware acceleration in firefox stopped it hanging at random times (thanks j95!)

The 314.14 seems to be getting the most positive results. Good luck.

Edit: Typo in the V# 314. 22