So my room mate RMA'd his card too, we tried uninstalling drivers and reinstalling older versions, along with disabling CCC but still got the same issue, also with the changing dvi ports, same thing didnt work , and lastly underclocking the gpu we had no luck. Sapphire just received my RMA yesterday. Im hoping this is like a 2 day process thing and i'll see my new or fixed gpu within a week timeframe. by any chance does anyone have the uber bios for download. My second r9 290x finally came in and im unsure whether to return it and get a 780ti or just xfire when i receive my rma card back..
btw, if you're overclocking with msia (which you most likely are), don't enable overclocking in CCC. leave it unchecked. and disable ulps and lower your vram clock. If you still get black screens, rma.
You need to understand that there are TWO causes of black screens. The first is emergency thermal shutdown and this is by design. If the GPU gets so hot that thermal throttling won't stop the temp rise (as shown by that battlelog post of 800-900 oscillating at 94C), the card will black screen and shut itself down. I was able to force a black screen in valley, by disabling LLC through IC commands in Afterburner. (don't EVER do that on air, EVER). Oh man did that card heat up....literally 5C a SECOND (80C, 85C, 92C, BAM). The second seems to be a fault somewhere, since some people have said this happens when browsing or just on the desktop, or within seconds/minutes of starting something like Heaven or Valley (long before the temps ever have a chance to get hot). And God knows what is causing that, but something is making the card shutdown. What were the temps in that Battlefield 4 recording where the screen was captured?
by any chance has anyone had exp with RMA's with Sapphire? was wondering how long am i looking for this rma process to be? (they received my card yesterday)
I've gotten 3 hard crashes in BF4 while loading, all of which froze my computer... Easily the worst PC game I've ever dealt with. I haven't read any of the replies here, but I'd say it's probably just the game, not the 290x.
So after doing a bunch of stuff suggested by postman and eclap I've not had a single crash since. Played bf4 last night and some more today ran some benches and everything seems ok. Just about to stick some on loop to get it cooking so we will see. Oh and my sound card is not in the machine... I also found a setting to put my pcix into gen3 mode... it still shows gen2 in gpuz but who knows!, I start narrowing it down after it passes the benches.
If GPUz shows PCIE 2.0 when the cards are under load, then they are at PCIe 2.0, and not 3.0. Also, don't get your hopes up so quickly, like I said, if your issue is gone for a few days, then great, but it's only gone for one day, keep crossing your fingers.
now that you disabled ulps, try putting the vram clocks back up. even overclcock if you wish. hopefully everything will stay fine. fingers crossed mate.
hmmm, could have been ULPS then. keep pushing the card mate, hopefully ULPS off eliminated your problem.
I noticed in bf4 memory clocks were less jumpy. Also mobo is still set to gen 3 pcix I know that doesn't magically make it run at gen3 but it could of been a compatibility error. Hell it could even be my sound card.
I'm pretty sure the issue will be back. I don't have a sound card and I do run PCIe 3.0 already and I also have those black screens.
yeah. I would put the sound card back in to see if it makes things worse. if not, it was either ULPS or PCI-E compatibility. if the crashing isn't occurring with the sound card back in, try enabling ULPS again. Most disable it though. it's proven to be troublesome.
I would not be so fast in returning into state at which it was failing. Let it run as you have it now for day or two. Give it good load. Then test with sound card 1st as it's priority 1 over ULPS, which we do not need. While you allow 100% fan speed you can try 90°C target instead of 95°C target. As measurement may not be exact and self lock temperature may be reached due to error in measurement.
We will see I guess fan is set to 100% that's one of the things I changed. I'm going to run another 24 hours of benches and games before I start trying to find out what I changed to fix it.
So what precisely have we changed to (maybe) get it working? Disabling ULPS via Afterburner (and that's it)? Or have we changed the fan speed too?
Or maybe nothing, and the card started working again for a while, just like it happened to me and many others where it started working out of nowhere. :3eyes:
Sure I've had this before, got 24 hours without a black screen once, solid Bioshock Infinite over a weekend, then Metro crashed in 10 minutes, tried BF4, 1 hour then a crash. But i'd like to try out the fix if possible. The people i bought my card from have given a single option. - Their RMA process, which may be 5-10 days, they did not elaborate on the testing procedure. I've linked them to all of the threads across the Internet, but they are not really taking it on board. If they don't find a problem (which i guess they wont), they send the card back to me. If they find a fault, i can have a refund or an exchange. So frustrating. (P.S Still within 30 days too, but the BF4 code means they wont accept it back as 'unwanted'.)