It's such a shame that AMD dropped the ball on the Fury X's overclocking potential. The GM200 (especially cut down as I assume the SMMs Nvidia slices are the slower ones even if they're not totally defective) can be a real screamer on OC.
The reference blower is a bit on the loud side. I used it for a couple days on my 980 TI to verify it works before ripping it off and throwing EK's Titan X waterblock and back plate on it. I've had it up to 1492MHz and below 40C load after 15 passes of Heaven 4.0. Such a beast of an overclocker that the Fury X doesn't even become an option at that point. If it had great overclocking ability with an included liquid cooler it would really be an attractive option (assuming they work on the overall smoothness through drivers).
Great comment, and I say that because I think the exact same way
I was looking forward to Fury X but got really annoyed by the lackluster OC potential. In my view, reviewing GPUs at stock speeds simply doesn't cut it.
Even if Fury X and 980 Ti were neck-in-neck at 1440p, you have to factor in the OC differential between the two. Hawaii was also a poor overclocker.
This didn't used to be the case for AMD, but these last few generations their GPUs have been terrabad at overclocking. Not good.
I have no need for this kind of power at the moment (1080p 55" TV, though DSR is beautiful) so I usually run it at stock, but I plan on getting at least one (maybe more) VR headsets when they release and I don't want to compromise there. I decided to see what it could do and I'm absolutely impressed. I have no doubt the Titan X and 980 TI will still be the defacto single GPU cars when SteamVR hits the streets.
Yeah, I'm either getting 980 or 980 Ti. I'm leaning towards 980 and for a simple reason: Pascal. I'm getting the Valve headset and not Oculus, but even so I'm giving it at least 6 months before we'll have any decent ecosystem(and realistically give it at least a year). So Pascal is supposed to be out in Q1 2016 but as usual, look for a late spring/summer release, so about a year from now.
Better to upgrade to Pascal than to go for 980 Ti when most of the VR content will not really be available until many months after you buy the GPU and by that time Pascal wil be out or almost out, so what was the point?
Like you, I don't want to make any compromises on VR either and will aim for big Pascal right off the bat. And giving the headset(s) a few months to get the initial patches/inevitable software issues ironed out is mandatory anyway.