I was using MSI Afterburner and set my RAM clock speed to 6300 on each 390. Honestly it was a typo...I meant to turn it up to 2300.
Again though, the VRAM...not the core clock. I left the core clock at 1050. And I didn't touch the voltage at all.
Also, I tried repairing Windows 7...it's freezing.
Welp, well reinstalling windows never hurt anyone.
Agreed. I was thinking it might not be fried since it is still outputting video without the drivers...it's only after I install the drivers that I have a problem.
so its not qdr? you clocked it over 600% up?
I don't know if it was or not.
I'm going to install Windows on a secondary drive and see if I have the same problems with a fresh installation. If I do, I know I'm screwed.
Time to RMA dem bois then.
A thing to note is that the power controller on 290s, 390s and Fury lineup (same controller) are really odd beasts, when adjusting the voltage with trixx my card (fury) tended to lock up at a set voltage, typically 0.9v.
Stayed like that regardless if I rebooted or not and the only thing that helped was keep rebooting and or pulling the card from the machine for a while, also flashing the VGA bios worked too.
Did so a good 3-4 times before I gave up with trixx, now AB supports Fiji voltage control.
Yes, I'm done with AMD forever. I'm buying a 980 TI and never looking back. These cards have been nothing but a headache since I bought them.
Nothing wrong with AMD though, and it obviously sounds like a headache because I'm doing stuff nobody should really do regardless.
Modifying the VGA bios isn't a great idea unless you know what you're getting into.
Just check if they boot fine on a fresh install, and test each card by itself.
Good luck overclocking that high nvidia cards. (you shouldn't touch oc, if you don't know what you're doing)
Doubt they will RMA though, as nvidia cards keep oc history in nvram - so they can decline rma.
Gents, listen...I've had BSODs and temps approaching 90 C on these cards playing GTA V @ 1080p, not even maxing it out at that. I've NEVER had a BSOD with an Nvidia card, and I'm not some goddamn fanboy either. You cannot tell me that a typo w/6ghz overclock on VRAM is what wrecked these POS cards. I know this to be true because I just ran a search and others have gotten stable OCs in the same range with 390s.
I've had 680 SLI and I've had 390 Crossfire. 390s are pure dogshit, and yes...I'll scream from the heavens that these are not 'nerd' cards or 'enthusiast' cards. These are cards that are for people looking to save a buck like I did, and I was fucking S T U P I D for doing so. NEVER again, I'll NEVER give them one more red cent. Because had I spent the $620 on a single 980 TI I wouldn't be bitching in this thread, I'd be playing my goddamn game.
Now, I'm going to put my order through on Amazon and work this weekend to pay for it. Later.
How on earth did they reach 90 C? And it isn't really dangerous for them to get that hot at all.
390 is a great card at its price, but getting two lower end cards instead of a single higher end one is a bad idea in of itself, a single 980ti/Fury X/nano/none X would do you much better.
Manufacturing quality there is no difference between Nvidia and AMD either, if anything the power delivery on high end AMD cards are much better than reference nvidia ones.
Just check if they play fine with a new windows install and if the don't just RMA them and get your money back and get another card.
1 Like
GTAV isn't even stable on consoles, that's hardly damning evidence.
All I got from this thread is that you killed your cards by trying to OC the RAM by 600%, and think it's AMD's fault for making shit cards. That alone is funny because what you actually bought are AIB cards which have little to no involvement from AMD besides the chip planted in the middle and the amount of RAM. Blame the company you bought them from, not AMD.
Good luck with the 980 Ti. Should provide wonderful performance until the month after Pascal launches.
hate to rub it in but that sure sounds like it !
I have been using ATI for over 15 years, at the time the war was between Matrox and ATI....(time does fly).
They have their flaws but calling them shit.......NOPE ! that they are not !
I'm going perhaps point out the obvious - you should not have to overclock the 390 to pull maximum performance from Shadow of Mordor. I suspect one of two things was going on:
1) You have a bottleneck at your CPU. I've had this issue and I've seen others have it. I had an old i7 that was the culprit.
2) You simply had a bad GPU, but more times than not I've seen #1 as the issue.
i believe your OC profile is still active somehow and that is what is causing the issue
try reinstalling afterburner and disabling the profile from afterburner
or disable it from the crimson drivers
It seems like you are done with trying to fix this but i would like to add that i agree with yuurei. It seems the OC profile is still active. If you reinstall both the drivers and afterburner then set both of them to stock settings it might help. Also as far as it running at 90 is concerned that is perfectly fine for a 390 its the same chip as the 290/290x and it was designed to run at 95 all the time when under load.
This is Slightly off topic However this reminds me of the old black screen crash issues that some 290's and 290x's had. Now i am something of an expert on that particular problem, thanks to my 290x being quite badly affected. long story short the only solution that gives me 100% stability (before doing this i would get the black screen about 3-5 times a day and i have not seen it since) is to set core voltage and aux voltage to +6, power limit to +1 and the memory clock to 1000. As i said i was badly affected, i have seen some affected cards work fine with the memory set to 1150, some even with no down clock. I have not heard of the same problem affecting 390's, but it might be worth a try.
proper 290x never ran above 80'C and kept within 68-74'C.
(there were too many bad cooling models of 290's, thus i assume people assumed it was designed to run like that)
reference models were running at 90'C+ that's when throttling occurred.