There have been some rumors circulating in recent days that AMD’s Bobcat project, a CPU made for the Atom/Nano market, might have been taken behind the wood shed for a little “project culling.”  According this very short but sweet post at The Inquirer, the deal is done.  Another post at EE Times claims that the project is dead as well.  

SAN JOSE, Calif. — Rumors are running rampant that Advanced Micro Devices Inc. (AMD) has dropped its ”Bobcat” processor project. The processor is expected to compete with Intel Corp.’s Atom.

Intel’s Atom processor is geared for the ultra mobile PC market. AMD denied the rumors that it is dropping what some call the Atom smasher project, but analysts are hearing rumblings on the grapevine.

”There are reports that AMD has decided to cancel its Bobcat project, which if true, we would view positively. Recall that Bobcat was aimed at competing with the Atom and targeted mainly for low power applications,” said Avi Cohen, managing partner at Avian Securities LLC, in a report.

”We had always insisted that AMD would be stretched too thin and can ill afford to continue development of Bobcat given its struggling financial situation and limited resources,” he said. ”Assuming that it’s able to offer a product, we have a hard time imagining how AMD would be able to compete effectively not only with Intel which is spending heavily on marketing the Atom, but also a slew of well entrenched ARM-based processors that are widely used in other portable applications like handsets, PND, and gaming.’


But, another perspective is on the topic was posted at Arstechnica that claims the project might not be cancelled after all.

That brings us to the present day and the current rumor that Bobcat is canceled. We call shenanigans. What little formal guidance AMD has actually given on the chip to date has only confirmed its existence; any confusion on that topic is due to speculation from the press, not the company’s statements. The idea that AMD would cancel its next-generation Atom competitor is even more ludicrous if Bobcat is a derivative of Bulldozer—why kill the smaller, simpler, and less-expensive part at such a late date? Doing so would imply that Bobcat consumed too much power to fit inside a netbook (bad), while simultaneously failing to meet a performance level that could make it attractive in a low-end notebook (worse). 

There’s no question that AMD’s roadmap is currently cloudy with respect to both Bulldozer and any 32nm transition; the company’s product guidance through 2010 doesn’t show an obvious point denoting when Sandtiger (Bulldozer for servers) would enter the market. One fact that shines through the murk, however, is that canceling Bobcat would mean things at AMD are an order of magnitude worse than anything we’ve seen thus far, to the point that the company’s roadmap through 2010 would seem less of a guidepost, and more of a cliff. 

Put simply, I think not.