That's great stuff, thanks for taking the time!
Thanks Sir!
I think it's impressive that their VP actually addressed the negative feedback from Fermi in the video. It's good to know that these companies actually take gamers' feedback seriously.
but yet this 680 is the holy grail of gpu's
I don't think GTX680 @ $500 is a holy grail of GPUs. I think this launch is nothing special in terms of generational performance leap from GTX580 but it signifies more than that. Since HD4870/5870/6970/7970, it's been assumed by a lot of people that AMD's engineers had better knowledge/skill in designing the most efficient GPUs for gaming. With Kepler, this notion is out of the window. It seems G80 (8800GTX), GT200 (GTX280), GTX480 (Fermi) were large die chips because of their GPGPU backbone and FP64 support functionality. It really took NV just a change of direction to come up with a very efficient sub-300mm^2 chip. I think this is going to provide NV with confidence that not only can they design a very efficient chip, but they may still be able to maintain GPGPU lead when they launch 500mm^2 "Big Kepler."
This launch is more of a signal to AMD that NV is able to make both lean and efficient GPUs and still have the option of going with a 500mm^2 monster given their historical track record of designing large die chips. It appears AMD's untouchable trifecta - performance/mm^2, performance/watt, performance/transistor - is about to vanish. And that's a BIG deal for AMD since their core competitive advantage not only on the desktop space but in the mobile space has been efficiency.
Essentially if NV can design a more efficient architecture and still maintain the ability to scale that same architecture to larger 500mm^2 die GPUs, then AMD is losing out its key competitive advantages in all market segments. I hope AMD responds to this with Enhanced GCN in 2013. I think they can because Pitcairn is already a better designed chip than Tahiti is.