No matter what i try, the difference is always 92 mhz in difference between the cards. Guess this is the reason why clocks will not go up further on this card.
Just found in GPUZ that the Power consumption on 1 of the cards is weird. So exactly the same.Īlso when i take the GPU1 offset higher, the gap (ofc) is closing between them. GPU-Z saying both cards are clock 1070, Boost 1149. How can the be such a big difference while the temp is still so low. Now when i run Heaven i find the following: GPU core clock from the 1st clock is 1163 between 50% and 98% load around 51☌Īnd the 2nd GPU core clock is 1267 between 50% and 97% load and temps around 54☌īios: both on F5 80.04.31.00.58 (P2004-2005)Īs said the setting are the same while in synced or manually set. With or without Sync on i did a start test with my old oc settings from when having 1 card. Something i found really weird is the following: Different products have overshot their power targets by different margins throughout the past.Been a while since i was on here, and after buying me 2nd Gigabyte GTX670 OC(GV-N670OC-2GD) i was trying to find max oc on both cards. Still, the whole "out of spec" deal tbh is way more tame than people make it out to be.
I could not give 2 flying fudges about whether perf/watt was better or worse than cards that aren't even in the same performance/price bracket. Honestly, the whole 6-pin connector was more gimmick "look what we can do with 6 pins instead of 8! *hint hint* *wink wink*". Edit: Just found in GPUZ that the Power consumption on 1 of the cards is weird. Also when i take the GPU1 offset higher, the gap (ofc) is closing between them. It is my understanding though that the 1266 boost clock speed on the 480 is pretty much fixed, so poorly binned chips may well overshoot the power target by quite a margin. GPU-Z saying both cards are GPU clock 1070, Boost 1149. So they more or less kept within the same power constraints because any efficiency deficit was made up by a lower boost clock. W1zzard however tell us something like lower better overclocking with regular cooling while higher better oc with better cooling. Last I checked gpu-zs ASIC quality check feature didnt even work for most Nvidia cards, while the number reported is useless for AMD cards. Like f.ex G1 Gaming 970 has power target A, thermal target B, ASIC score C, final boost clock D. Take 100 reference 580 boards, or 100 reference 7970 boards as an example. It was a combination of metrics that gave rise to a final boost clock which was predetermined.
Contrary to what people believed about GPU boost on the NV side of things to begin with, boost clocks weren't random. This could be, but I wonder if this is possibly part of the issues people are having with the card drawing too much power over the PCIe bus.Ĭould very well be.
0 kommentar(er)
