Furmark is nothing but power hog. Its not really a stability benchmark because driver Over current protection limit kicks in sooner before you find stability issues.. This started since GF400 or 500 series and its build in every gpu since then.
I get the same voltages and clocks in furmark as I do in Witcher 3, tend to have slightly less in Heaven. So wouldn't furmark be good for testing stability? If I'm getting a crash in furmark wouldn't that be a good sign that things aren't stable? Not really understanding.
No the memory is the exact same thickness as the other pads. 0.5mm It's really important that you don't have the card bending like that. Not only because the PCB is bending, but I can guarantee some of (if not all) the thinner pads aren't making contact which is BAD. They gave you the wrong sized pads because my EK block only came with 0.5mm pads. No other thickness and as you see it works perfectly.
Hmm, I can't seem to change the voltage on my cards through AB either. Stuck at 1.19 no matter where the slider is.
So here's a video of the fan whine I'm getting. It gets really bad at 40 seconds. Think I should send this back to newegg for a replacement? https://www.youtube.com/watch?v=Bslf_lJ4OPM
Well Im speaking of my own experience owning GF110 & GK110 chips, it either throttled down to base clock - GK110 or lowered fps - GF110 when it hit certain power limit. Imo Resident Evil6, 3dmark Vantage - extreme, 3dmark11 - extreme, 3dmark Skydiver - especially demo scene, Vally, HEaven, FFXIV heaven, Monster Hunter online, Sniper Elite2, HqardReset - max with 4xmsaa benchmark and few more are all those real stability indicators, especially 1st four. Furmark is just like for cpu Intel Burn test, it overloads/overheats cpu for nothing.. E.g. what helps if it fails later in a simple game like BF4 or X264 encoding/decoding.
I've yet to actually have any instability in 3DMark in situations that I did have problems in games, furmark or Heaven. I just think a good spread of things is a good way to test and furmark stresses things differently so it's a good one to throw in there.
Does it always happen around the same fan speed regardless of load or only when you're maxing out your cards?
FYI - 1484 was my max before modding my bios, and I was already using an aggressive fan profile. Upping the voltage didn't help me because I was hitting the 110% power limit which was causing my clocks to throttle. It's pretty damn shameful that review units allowed the power to go up to 120% which in turn allowed for higher overclocks.
Thanks for confirming. Yea this is really bad that they supplied the incorrect thermal pads. Lucky i still have my old 290x waterblocks and used the pads from them. Looks much better now. :thumbup: Installed .5mm Pads
Anyone else switched from 970 SLI to 980Ti and can tell me the difference while gaming in 1440p? Very nice! But don't they sweat like hell in that H440?
I can tell a big difference Netherwind. Smoother for sure and frame rate is as good in SLI games but still smoother. Games that don't support SLI are double the fps.
I think 970 SLI is actually a bit faster than a 980TI. You'll only notice a difference in games that require huge amounts of VRAM (possibly at 4K) or games that don't support SLI.
I remembered that someone had written this HonoredShadow but I forgot it was you It's still some ~200€ I have to invest in this but I'd really like to do it. It's very hard to find actual benchmarks, I've just seen one on youtube and according to them 970 SLI is supposed to be a bit faster. I'm playing at 1440p but I only think that ACU, FC4 and Mordor punched the 3,5GB "limit". Btw, nice OC, I take it you got lucky in the silicone lottery.
Yup. But I've had to turn it down a little bit after a 5120x2880 torture test in The Witcher 3. Not a very common scenario, but my OC proved unstable. So I tuned it down to 1520/7840, which seems to survive even that amount of torture. And that's with +33mV from afterburner. 79.6% ASIC quality. Not bad, all things considered.