This is what I mean: Because nVidia is releasing information on THEIR timetable, they are being attacked and written off for dead based on "leaks" from sites with a clear ATI biases and/or funding. Here's how the lose-lose works:There's no "what if." The leaked benchmarks are already telling of Fermi's performance, as is Nvidia's unwillingness to share numbers.
1. nVidia DOESN'T put out numbers way in advance and various sites and fanboys call this proof that they've failed. "If it was so good, they'd be bragging about it."
B. nVidia DOES put out benchmark numbers way in advance and various sites and fanboys howl that since real cards aren't in neutral hands for testing, this means that it's all vaporware and proof that they've failed. "That's just what they say and they're using a benchmark specifically tailored to the one thing that doesn't totally suck on this hot, loud, power-hungry, late hunk of FAIL!!!"
We'll know tomorrow (or Monday) just what Fermi is or isn't. That so many are constructing arguments designed to make nVidia look like losers no matter what - "Only 50% faster than my 5870 for the same price? FAIL!!! Gonna buy another for CF. Buh-bye, green losers." - shows more than support for the current superior cards, but definite animus against the "other guys."
It reminds me of when 3Dfx fanboys used to retort to every new nVidia product with sneers of, "Oh, yeah? Well, Voodoo [next number higher] is gonna murder nVidia! You don't need 32-bit rendering! It's just marketing hype since no games use it; same for hardware T&L! Voodoo rules!!!" How'd that work out for 3Dfx anyway?
In a day or four, we'll know what's reality and what's FUD. Patience.