So I just got my 7970 DCu ii TOP and I wanted to put up some feed back. First game I started playing was BF3, played for about 10 - 15 minutes. I was running HWinfo64 to check out how things were looking stock out of the box, and it is causing me a lot of grief right out of the box. Not over clocking, running Catalyst 12.3 drivers. I grabbed this screen shot when the VRM hit 106 C. Shortly there after, my entire screen artifacted and then locked up my PC. I rebooted, and mid way through windows loading, locked up again with an artifacted screen. Following that, I tried to power up, and even my bios screen was artifacting, and windows would BSOD mid way through boot up. So I powered down, powered off the psu, and let it sit for 20 minutes. Repeated boot up same issue. I took the card out, looked at it, seems ok, screws on heatsink are not loose, heat sink didn't look damanged. I put it back in, powered up, everything was fine. Fired up BF3, for about 10 - 15 minutes, VRM temp shot up to 105 C+, artifacting, locked up. Same issue, couldn't boot up at all because of artifacting bios screen then bsod on windows load. So I took the GPU, reseated it, boot up, everything is fine.. PC runs no problem with out load, so I don't think the card is loose or something, I even ran the fans at 100% for 20+ minutes to make sure it isn't unseating itself. After some more playing around on stock clocks, my opinion at this point is that there seems to be a combination of design flaws. The first being, the VRM's run hot - it seems reference cards run in the 60's, the VRM seems to be designed to run hotter on the DC2T card. This comes into what I think is the second design flaw. With the stock cooling profile, my gpu never gets hot enough to bring the card past 10% fan speed. This results in low air flow, and the VRM heating up into the 105c + temps. I modified the profile to kick the fan up to 50% when the gpu hits 50c, and my VRM never went above 89c after 2hrs of gaming BF3. I haven't had any time to play with the profile of lower speed and VRM temps. This leads to a third issue - since the VRMs seem to be designed to run hot, and require manual fan speed profiling to keep them in a more respectable temp, it means if you want to OC this card, your VRM's are going to have much less head room before getting really hot, even with a high fan speed. So the card seems to be design too well in the cooling department for the GPU, but forgot about the VRM cooling which has put a ceiling on OC'n the card. At this point, I am considering sending the card back. I was really hoping to run this at a super low fan speed to keep it dead quiet. I also want all the additional ports for screens so its a tough call. OC'n the card would be nice too, but I think with out some better designed cooling, the VRM temps are going to be an issue which is unfortunate since the GPU cooling is so damn good. Final thoughts; I can't decide whether to keep it, and if they issue a newer revision then to RMA it at a later date, or send it back now and wait with nothing in my rig and not gaming until there is a replacement I want. I'd really like to have the dual DVI and 4 DP's on a card, and something that is super silent. The Gigabyte Windowforce card would be my next pick but I kind of want the option of not having to use adapters for the out put. Asus you disappoint me on a $700 GPU
Wait, so you are getting artifacts and BSOD and you are wondering if you should send back or not? Maybe I missed something but I'd RMA instantly, something is wrong with the card if you ask me. Your set up looks sweet though, what case is that?
Mmm... nice case. That cable management is very Antec.... as is the partition from the motherboard to powersupply... not quite right to be the P280 Antec I initially thought.... What case is that?! Nice to see that you mean business! Oh, and your card... sounds dead to me. Thats just too hot. Its gonna melt right through the bottom of that pretty case! Try getting it replaced with another of the same card if the retailer is willing to play ball and you just HAVE to have all those outputs. Keep me updated. Im super intrigued as a potential buyer of one of these GPU's, as well.
ya I am RMA'n it today back to NCIX. The thing is the card runs great if you over ride the fan and set it up to a higher temp. There is a massive thread on HardOCP about these having really high VRM temps, but nothing about BSOD and lock ups, so your right i'll send it back. The case is a Fractal Design Define XL. Very similar to the HAF-X inside, but the outside is ultra clean, and the case is sound proofed.
I was about to call HAF-X before you posted that picture, there was no way in hell that was Antec quality. I'd RMA it ASAP, esp. if the VRM temps are sky rocketing that high. And you said you checked everything (voltages, fan speed %, etc) so its got to be something else with the card.
Got lucky - 8 more came in stock today so I queued up a cross ship Plus a couple days to test CFX before sending in the RMA heheheh Gonna be a tight fit
Not going to test well in CF if the original card can't even run in single mode w/o crashing due to heat. Good luck though..
Hell Id try it to if for nothing else than the badassery of having two of the top video cards on the market in my comp. Never had anything like that before... Ah, and the reason I called Antec was the cable management. Its extremely similar to one of their P series towers. It also has the same hard drive and powersupply arrangement. But it in no way had the same attention to detail and sound dampening. That really is a spectacular case you have there! Congrats on the awesome system.