Tegra = typical Nvidia-story
Tegra = loud and bold promises, history of wasting money on two generations while promising everything for OEMs then not delivering half of it and then being rejected everywhere.
I wonder how long they will be able to fool investors - the Fermi-blunder was a hit on the head for many especially when it turned out the CEO was actually showing them a plastic mockup instead of a prototype card last year this time... indeed, the card came out ~8 months later. Oops.
Tegra is even worse: they probably spent billions on it and there's absolutely no market share to show; if it does not change in the next 2-3 quarters investors will need another fairy tale instead of Tegra.
Ahh and claiming Fusion and Sandy Bridge won't effect their market is downright hilarious: they will hit first exactly in their supposed Tegra-aimed market, the mobile/notebook market. On the other hand if we consider they dont' have Tegra market share it's true... :)
One thing is sure, the faster Fusion or Sandy Bridge gets adopted the faster Nvidia's results will worsen - being short of x86 license they really need a something big to maintain their positions on the mind- and long term and I don't see their Fermi-based market booming in such pace anytime soon.