Yes, but that issue is part of what makes the 980 Ti not as much of a success as it could be, perhaps. There are some prevailing issues on either side of this context, both with AMD and NVidia. Yet to me these are two totally different companies with totally different business models and objectives. AMD is about pushing hardware design and innovation; they are a silicon technology house. NVidia, though, is a software company that also sells hardware, as so proudly claimed by their CEO. So, I think it's proper to be a bit more critical of NVidia and their software, while with AMD, you simply have to expect driver issues as part of ownership, because software is secondary to their main objectives.
True enough, in the greater scheme Nvidia's TDR issue is a minus, but as far as I'm concerned, any company that puts out product should support it fully, and represent it accurately. You don't get a pass because you're known to be weak in an area of product support any more than Nvidia should get a pass for how it high-handedly managed the GTX 970 3.5+0.5GB issue because their hierarchy is known for its arrogance.
And if we hold Nvidia to a higher standard on software because it makes up a substantial part of their products, do we not hold AMD's hardware to the same higher standard? There was considerable uproar over Nvidia's capping voltage limits for their Kepler/Maxwell cards, so how much more scrutiny should AMD be under for locking it down
completely? Moreover if this is just a
temporary measure, why not allow it from the start? Board partners had access to the card for some time, yet are totally unable to work voltage control into their OC utilities? A sceptic might surmise that allowing voltage control for OC'ing might just have impacted on reviews negatively. Overclocking doesn't seem to yield high real world returns, but I'm pretty sure turning up the voltage control would add more power, increase heat, and in turn, require that Gentle Typhoon fan to spin a little faster than the 1500rpm is was sent out with.
To run it back on topic, this makes FuryX a grand success, and the 980 Ti a failure?
Ha, if it were only that simple. A success is usually defined for a consumer product by sales, feedback from both the industry and consumer, and whether the product realized all- or most of its goals. By all accounts the Fury's launch was delayed considerably by performance issues with
software. If the company had launched the card in a timely manner when the hardware was ready, and pre-empted the 980 Ti (and possibly the Titan X) which would have presented a completely different picture of the card to both the masses and consumers. Fury X with no GM 200 competition would have been raved about in every public forum imaginable. The card now suffers not from its own image, but that by comparison with the competition. Would AMD judge what would have been a PR bonanza, but is now "nice try" territory a success? Sales are brisk, but I doubt that there is any significant volume to speak of, and how many people who held their breaths waiting for Fury X, exhaled when it dropped and opted to either sit it out or buy a 980 Ti ?
So, regarding your ( I'm presuming largely rhetorical) question, I would point you towards the
TPU poll, and ask if the voting would be the same if the 980 Ti option was replaced with just the Titan X at $1K or with the GTX 980 in its place?
So it probably requires a definition of success. Yes the card is selling, but there also seems to be a general air of "meh" and deflation. I'm betting that AMD banked on a little more than "meh" for their largest GPU in their history by a considerable margin utilizing a revolutionary (discounting that Intel got 2.5D stacked memory into product some time ago) memory technology. Personally it is a storm in a teacup. The card isn't bad, but from my viewpoint it isn't great either. Reference only, voltage locked doesn't scream "BUY ME!" for the enthusiast tinkerer in me. In the greater scheme, I suspect that the card will sell, and will continue to do so when AMD institute their inevitable price cuts, and in a year we'll be debating the veracity of Arctic Islands and Pascal benchmarks popping up on gossip sites reprinting some Chinese adolescents first attempt at making a bar chart in Excel.
