Why keep saying it like that, you're trying to hype it up. No graphics card has ever been 3 times faster than an older one, ever. It's not physically possible. At best it will have twice the transistors, that is a physical limitation. The reality is that they are unlikely to make 600mm^2 cores, AMD or Nvidia, on 16nm. If Nvidia goes big on 16nm it will unlikely go much beyond 500mm^2, more realistically it will do a midsized core like 680gtx then might do a bigger part but well below 600mm^2.
Nvidia also is likely to put back in compute and hardware scheduler making for a smaller than most gens increase in transistors for direct end gaming performance.
There is categorically no chance of 5x the speed, none. At best it will be 2x Titan X for a similarly sized core, maybe 2.5x in very specific games which overuse certain features, but would be no where near that with said feature turned off. IE Cod 56 might be 60fps on Titan X and 100fps on Titan Y with the same settings, but 30fps vs 80fps with some new feature turned on(still not very likely).
Most gens you get lucky to get 80% more performance out of twice the transistors, 65-75% is more likely and Pascal has very little chance of either of those because of the tiny chance of making another 600mm^2 core early on 16nm and having to add those other things back in.
I'd say a 600mm^2 core with compute/scheduler back would struggle to be 60% faster, but I very much doubt they'll get a core above 400mm^2 in 2016, which will struggle to be 25% faster.
the silence is annoying indeed, I hope it ends soon. Like the other guy said, maybe there won't be a Ti and they just move on to Volta. There didn't do one for the 4, 5 and 6 series, but did for the 2, 7 and 9. They can do whatever the heck they want and need in the end. Whatever comes out, it'll be better than my current card 
and if only the price was right... How long does it usually takes for inflated new GPU prices to come down?
I am genuinely concerned that performance will not be there on "lightweight" Pascal.
