LOL, aye spider solitaire pushes the system! Proving a point to the non-believers mate! Was getting dogs abuse! Regarding my gaming/full GPU load temps!
I'm moving to a new place so I've not had time to try this out just yet. Will try some stuff including repair when I got internet and have set up my computer. Didn't know they released a new BIOS so I'll try it out. Can tell you though that this happens from the moment I launch the game until I turn it off. And the drops occur almost exactly every second. Not sure about GPU Power or GPU Load, I'll check that out too.
Those are nice cards bro, though I'm getting red/blue dots crashes sometimes, that ever happened to you? Only happens during quite high load. The bsod that comes up is driver irql not less or equal (ntoskrnl.exe). Playing d3 @ 4k dsr, (STABLE) overclocked it can crash, just red dots all over the screen, I read a bit on the net, says it's either too high load or card is faulty. But regardles of load, it should still ****ing FUNCTION right, it shouldn't bsod with red dots all over lol. If it's a core clock instability the screen just goes black/freezes and the driver resets itself, not dots around with reds or blues. If it's a mem oc there's just artifacts. So this is by no means an oc instability issue. This is something else.. maybe something to do with power, it just doesn't handle the load or whatever happens. I played with stable clocks, red dots.... lowered the oc even more, far from being unstable, still red dots... playing @ stock, no crash..... But still, it shouldn't crash/bsod this way... even if the load is high... I think the card is faulty tbh. No matter the load it should still handle it... cba to send to RMA, **** it.. Will see how it goes... Any thoughts on this? The dude I bought the card from told me to test in another pc but I cba doing that, I don't have any other pc lol, + my old 7950s worked fine, no issue there, even though they were both oced to hell. I'm gonna look at the dump once more and see what it says, hopefully it should give more insight.
Jonny, in all honesty, you haven't ever posted your load temps anywhere. MadMax, is not and never will be used by any serious person to assess GPU load temps, not even at 1440p and 120hz I bet it wasn't putting the GPU under full sustained load The screenshots you posted are also meaningless in all honesty. They dont show anywhere (in graph form) periods of full GPU load and corresponding temps. Also who measures a GPUs cooler effectivness in "average temp" ? wtf ? People want to know peak temps eg worst case, so in your screenshots I can see that was 57'c from your gaming session in MadMax. Nice temp, but we dont know if your card was fully loaded for sustained period of time in that game. Now i'm glad you love your card and are pleased with it and its got a very good cooler on it, well for an air cooler anyway But lets not try and kid anyone into thinking anything else when we can all see reviews of this card and know that its load temps are late 68-72'c range when fully stressed. It probably looks like this is turning into abit of a Witch hunt at the minute, that isnt the intention, its just that people (quite rightly so) will jump all over 'questionable' claims around here. Why wouldn't they, its an enthusiast site. All the best P.S If you loaded up a benchmark such as Heaven or Valley, maxed out all the settings, then let it run for 30mins or more with a GPU time graph monitoring in the background eg nVidiaInspector sensor monitoing or Afterburner osd...then posted a screenshot of that.....jobs a good'un...
Will do! I will run heaven to give you a better idea of load temps. Also it was CPU average temps not GPU! It was max and min temps for the GPU, give me some credit! Mad Max btw with peaking at 98% as vsync was off.
Then there is something that doesn't exactly run right on your system. The part where it starts stuttering from start to end... it must be something else. Something not running right in the system. I would start with updating mobo bios, reverting all clocks to stock, then see how it works.
Ok dude no worries. Here is what I mean. From the graph you can see that GPU was loaded fully at 99% (bar the dips where the scene changes) untill I alt - tabbed out to take the screenie. GPU clock was solid at 1501mhz, voltage solid at 1193mV etc etc GPU Temp, the max value is what we are interested in. Mine maxed out at 50'c. This is what I would say was my load temp (not the current temp of 39'c not the low of 36'c during the benchmark, or some average of them all) The max temp it hit, the worst case scenario is what people want to compare !!!
Indeed bro. What I'll try when I'm in place is to first update the BIOS, then try at stock clocks (since everything will be at default), then try the 970s (now that SLI has been fixed on Z170).
470 artifacted, 460 stable, 450 stable, 450 is what I've been running with, since the crashes I've tried 440... 430.. but still crashes, no difference, don't think it's that. But i'll try running with 400 mem. What can I change the bios with, any maxwell bios editor? I wanna change power limit to 120, I wonder if it'll make a difference.... any thoughts?
dots are usually too high vram related. I got mine at 120 pt. helps with throttling. post 38 in this thread. http://forums.guru3d.com/showthread.php?t=394091&page=2 I just run the vrams +250.
Yeah because I see power going over 110 sometimes, I see in msi ab, says 113 max lol. wtf, 110 is max? So yeah.. maybe changing that pt helps.... files no longer available btw on those links it's fine, downloading now........ Hmm, I wonder, if I flash, would palit ever know in the hardware? If there would be any history of this or whatever.... if I do rma then I'll just flash back... but, could they know somehow I've flashed it before? ...
I doubt it, very easy to flash it back anyways. Been flashing all my cards since 680 sli without problems. 2 members here bough my 680`s already flashed and they are still going.
My 670's have been flashed since I bought them pretty much. They do get real close to that 150% pt I have though.
You need to save your stock bios with gpuz and put it in a safe place. I can talk you threw the rest of it if needed fairly easy.
Well I had the Overclocking itch over the weekend and decided to take a look at some custom BIOS for the GTX980Ti over at overclock.net. I ended up getting the 450w 1.28v max BIOS and mines reporting at 1.274 in AB but with the most minimal of core boost to 1232 (previously ran it at 1200) I'm hitting over 1500mhz on the core instantly where I was stuck at about 1430 +/- before. I've yet to give it more juice but i do know that when it hits 1600mhz on the boost frequency it fails pretty quickly so i think I'm hitting close to the maximum this chip will do but at only 46'C at 99% load with these clocks is good enough for me :banana:. More tweaking to come :nerd:
OK guys, some proper benchmark results for my Zotac Amp Extreme! Ran Heaven for 45 minutes! Core 1270Mhz VRAM 2000MHz (8GHz effective) Boost 1459MHz, Power boost 111%, stock voltage.
nah ive flashed before its fine guys, my 7950s could feel that quite much lol, made custom bioses with different max clocks and pt. i'll do this and report back
Loving my Zotac GTX 980 Ti Amp Extreme so far. One of the fastest non-ref cards, does 1500mhz core, and 75.7 asic!
the card didn't crash yesterday whilst playing d3 @ 4k dsr, played with 105 core, 400 mem (this time), 110 pt and 0 volt worked fine.... I wonder if mem could be the issue? will test more....