01-23-2014
07:04 AM
- last edited on
03-05-2024
08:01 PM
by
ROGBot
02-11-2015 07:00 PM
02-12-2015 02:08 AM
Ishan wrote:First of all, I wish you to see your card stable in all applications. You've already tested it on wide list. However I think all the burners like furmark burns only fillrate with reading a lot of texture data from vram, but do not write a lot. But 1000/1500 should be stable in all the cases.
Btw 3D Mark, Unigine, Firestorm etc never once showed me any artifacts.
That didn't work when done manually.Did you mean with other bios? I've found that during PowerPlay state change sometimes actual clocks set up in Afterburner or TRIXX does not apply instantly, sometimes the clocks from bios was set for a short period. So I think flashing the clocks same or lower than ones you're testing is important to get clean results.
02-12-2015 04:19 AM
dotachin wrote:
First of all, I wish you to see your card stable in all applications. You've already tested it on wide list. However I think all the burners like furmark burns only fillrate with reading a lot of texture data from vram, but do not write a lot. But 1000/1500 should be stable in all the cases.
dotachin wrote:
Did you mean with other bios? I've found that during PowerPlay state change sometimes actual clocks set up in Afterburner or TRIXX does not apply instantly, sometimes the clocks from bios was set for a short period. So I think flashing the clocks same or lower than ones you're testing is important to get clean results.
dotachin wrote:
I see you're familiar with tuning tools like Afterburner and ASUS GPU Tweak, can you please do some overclock experiments with AC:Unity when it will be downloaded, can you try 1100/1500? It would be very interesting to get the results of such tests from other card to better understand the conditions of artifacting and confirm or decline the theory that high core clocks are required to load vram enough to make the artifacts appear.
02-12-2015 04:40 AM
Ishan wrote:It would be the ultimate test solution, but it does not use more than 2GB of VRAM, so even if it will not show any errors, we cannot be sure the card is stable in checked profile. Same thing with 3DMark. So I've ended up with ACU and even removed intro vidoes to make it start faster.
Yeah i think there's no other vigorous test than OCCT
02-12-2015 04:49 AM
I meant manual settings done via GPU tweaks on the original vBIOS of the DC2T card.Oh, you meant you've tried to flash clocks in DC2T bios but got same artifacts. Then there should be some difference in timings or something else (but what could it be in?). Unfortunately, compare DC2 and DC2T bioses does not give any acceptable results, as almost half of DC2 bios is probably shifted or replaced in compare with DC2T. But it's still possible to compare the timings, i'll try to find them in it or find and compare original and optimized by The Stilt versions.
02-12-2015 05:10 AM
02-14-2015 06:22 AM
02-17-2015 11:11 PM
dotachin wrote:
Here's the bios with raised up timings to try for anyone who wants to experiment. With this bios card performs about 5% slower on same clocks, but helps to reach higher memory clocks without artifacting and without having to lower core clocks (helper to get +100MHz more on memory in my case, but may help better for others). The changed timings are only those altered by The Stilt in his optimizations.
02-18-2015 09:05 AM
Classysaurus wrote:For my card it gives +100 on mem without artifacts but this may vary for each card. But there's 5% performance loss (compared on same clocks) which should be same for all cards.
What kind of success have you experienced with this bios?