AMD pooh-poohs 'Atom smasher' cancellation claim
Why drop 'Bobcat' long before it's needed?
So farewell then, Bobcat, AMD's answer to Intel's 'Silverthorne' Atom processor family? AMD's denying it, but at least one analyst is tolling the low-power chip's death knell.
Avi Cohen, managing partner at US investment house Avian Securities, this week told clients: "There are reports that AMD has decided to cancel its Bobcat project."
Cohen regards this as a good thing. Trying to compete with Atom would be very hard, he reckons, severely limiting what AMD will get out of its investment in development and production resources for the processor.
To be fair, Cohen admits that he doesn't know for sure that the Bobcat project is over - his analysis is clearly conducted on the assumption that the rumours are true, but he provides no evidence that they might be.
For its part, AMD roundly rejected the speculation. ''Bobcat remains under development as an important future core for AMD," a spokesman told EETimes.
Talk of Bobcat's demise may arise from AMD's limited discussion of the part. While Bobcat's been known to be in development for some time, AMD has said little about it. However, in July, CEO Dirk Meyer said that AMD's response to Atom would be detailed in Q4.
It was assumed - by us, as well as other observers - that he meant Bobcat. In fact, he appears to have been referring to AMD's Ultra-Value Client (UVC) platform, which is based on a couple of Athlon processors, one single-core the other a dualie. AMD roadmaps show UVC being launched right at the time Meyer mentioned.
That, in turn, may well have persuaded some folk that AMD's dropping Bobcat for Athlon. But it's worth remembering that Intel has two strands of Atom: 'Diamondville' for mini laptops and 'Silverthorne' for handheld mobile internet devices (MIDs). UVC will go up against Diamondville, while Bobcat will target Silverthorne.
Right now, UVC is more important to AMD than Bobcat - the market for MIDs is simply too small, hence Cohen's comments. But, as ABI Research this week forecast, MIDs could account for 36m mobile device shipments in 2013 - 136m out of 400m - and that includes not only netbooks but regular laptops too.
Now's not the time for Bobcat - AMD has a few more years to perfect the product then release it just before world+dog goes MID mad. If, of course, it does...
Sponsored: DevOps and continuous delivery