Replacing stock TIM on GTX 470/480 Thread

Hi guys,
have been following the discussion about improving the cooling on 480/470 cards and would like to present a theory of my own. It stuck me that the exhaust aperture on the the rear of 470's is quite small, much smaller than the fan intake, and, the "grill" actually blocks alot of the available area, about 1/3 I would guess. So I cut them out :)
Here are some before and after figures;

GPU temp fan speed/%
92 2890/67%
90 2630/63%

now I know the change is slight, but lower fan speed for the same cooling effect indicates to me that the restricted airflow is the main limiting factor in the reference design of GTX 470.
This might go some way to explaining why changing TIM or providing better cool air to the card dont have much effect, they simply aren't where the issue is!
kind regards
 
Hi guys,
have been following the discussion about improving the cooling on 480/470 cards and would like to present a theory of my own. It stuck me that the exhaust aperture on the the rear of 470's is quite small, much smaller than the fan intake, and, the "grill" actually blocks alot of the available area, about 1/3 I would guess. So I cut them out :)
Here are some before and after figures;

GPU temp fan speed/%
92 2890/67%
90 2630/63%

now I know the change is slight, but lower fan speed for the same cooling effect indicates to me that the restricted airflow is the main limiting factor in the reference design of GTX 470.
This might go some way to explaining why changing TIM or providing better cool air to the card dont have much effect, they simply aren't where the issue is!
kind regards

Yep you're right. http://www.evga.com/forums/tm.aspx?m=377056
Hopefully the new brackets will be available separately soon.
 
So, combined with the tim replacement, this should produce a huge temp.drop. jk:rolleyes:
I've also been replacing cpu/gpu tim for years. I find the exposed heatpipes of the 470/480 heatsink with it's grooves and uneven surface to be the biggest problem when trying to find a suitable tim application method.
Used MX-3 and tried spreading it fully across the gpu, as in some EVGA guide, while putting a light "haze" across the heatsink. This resulted in lousy contact. Wound up putting five thick lines across the gpu, that resulted in better contact and coverage. As for temp. drops, haven't noticed any yet:p
 
Last edited:
This is a waste of time. Unless there was a mistake at the factory, there is no point in replacing the TIM. I did this on three GTX 470s with either Freeze or Ceramique using all three methods and there was no difference in temps, except that the single dot method resulted in higher idle temps.
 
I'm still working on it.
Any word yet? Because I'm ready to take this thing back to Fry's in the morning. The $ savings just isn't worth the lackluster warranty for me. Besides, EVGA's giving away like 18 games anyway, so that should more than make up for the difference. I did like the cooler though...has a death star look to it. :D

EDIT: And the packaging. EVGA could learn a thing or two from Galaxy's foam shell and thick cardboard box. I hate those flimsy little plastic wrap/thin cardboard doodads EVGA does.
 
Lack of better equipment, lol. Id trade setups in a heartbeat. :)
 
Back
Top