I agree perfectly well about "good enough" performance. But...
Right now, on pricewatch, you can find a new toshiba K6-2 380 12" screen notebook for $1k. That should be good enough too, for most things. And that's a preshrink K6-2, no doubt, the production costs and power consumption should be lower on the .18 micron line. The Transmeta PC-class chip is supposed to be available in 6 months, meaning maybe not on the street in notebooks or equivalent for a year. If somebody wants to make cheap notebooks, instead of following the the old PC model of faster, better, more for the same (high) price, we could maybe see $500 notebooks in a year. Or maybe not, a big piece of the price is the display, but a webpad type appliance is not going to undercut a notebook there either.
Will Transmeta be enough better than a K6-2 350/400 to beat that competition? Mobile parts are different, and don't show up on pricewatch or but a desktop K6-2 400 cpu can be had, quantity 1, for <$40, or $52 with shipping. I would guess OEMs aren't paying much more that $50 for the mobile K6-2 in quantity. Low power is nice, but is low power alone enough? The big rap against K6 was that it couldn't match Intel's floating point, which is another "good enough" issue. I wouldn't hold my breath on Transmeta's fp either.
I don't know, I'm all for innovation, and it's nice to see some new ideas in the CPU market, Intel has had a somewhat stifling effect. But I'm not sure, from what's come out so far, that Transmeta has real good odds here.
Cheers, Dan. |
