For the longest time, I had the hotter 7970 as the top card. On idle, it would be 10c hotter than bottom, and on load, it would be almost 20c hotter. I'm looking at 85c on top card and 65c on bottom card. At first, I thought this was normal because of the restricted airflow from the top card, however, I remember the gap between my two 570s in SLI was only about 10c, not 20c.
I decided to switch the bottom card with the top card to see what temps I would get, and lo and behold, the temp gap between the two minimized. At idle, they are almost the same, with the top card 5c hotter. On load, they would almost be identical in temperatures while they are around the 65-70c. It is until I put more game time in BF3, the top card would push out ahead and hit 80c and bottom would hit 70-75c, so the gap got minimized as well on load.
Obviously I made a good switch by putting the top card on the bottom and cooler card on top, but what contributes to the added heat of "hotter" cards? Bad manufacturing TIM application? I've never opened up the cooler and I don't want to mess with anything that I don't have to. Both cards are 86% ASIC quality with the same voltage applied. Is the only culprit bad TIM? or it's just some chips run hotter than others, even at same voltages, clocks, etc.
Or should I just leave it be? Since it's not really THAT hot, it's just hotter than the other card.
I decided to switch the bottom card with the top card to see what temps I would get, and lo and behold, the temp gap between the two minimized. At idle, they are almost the same, with the top card 5c hotter. On load, they would almost be identical in temperatures while they are around the 65-70c. It is until I put more game time in BF3, the top card would push out ahead and hit 80c and bottom would hit 70-75c, so the gap got minimized as well on load.
Obviously I made a good switch by putting the top card on the bottom and cooler card on top, but what contributes to the added heat of "hotter" cards? Bad manufacturing TIM application? I've never opened up the cooler and I don't want to mess with anything that I don't have to. Both cards are 86% ASIC quality with the same voltage applied. Is the only culprit bad TIM? or it's just some chips run hotter than others, even at same voltages, clocks, etc.
Or should I just leave it be? Since it's not really THAT hot, it's just hotter than the other card.
My BF3 Rig
(15 items) |
|
My BF3 Rig
(15 items) |
