CHIP DESIGNER AMD has ruled out making a move in the smartphone market, preferring to concentrate on tablets.
Rick Bergman, SVP and GM of AMD's products group told a conference that the chip designer has no plans to get into the smartphone market, saying that its expertise in graphics does not suit that market. Instead it will be up to AMD's Z-series embedded chip to push X86 into the tablet market.
Bergman told the crowd "We haven't announced any plans to go in that handheld space. We've got plenty of opportunities... in server, notebook and now tablets, that's our immediate focus. But if the right circumstances come up and we can see a way to impact the market, we'll obviously continue to look."
AMD's decision to leave the smartphone market alone is an interesting one. It once again raises the question of whether X86 system-on-chip (SoC) designs are profitable.
Intel is trying extremely hard to put its Atom processor into smartphones and believes it can do so, but realistically it will at least need to use its 22nm Tri-gate process node in order to bring power consumption down to the levels that are needed in the embedded market. And even that might not be enough, as Intel has publicly said it will need to go beyond Moore's Law for its Atom processor, which really means it needs a very mature 22nm process node or more likely 14nm scale to really hit the sub-1W thermal design power required.
AMD relies on others to punch out its chip designs, and we have heard a few stories that some fabs are having trouble with next generation process nodes. Our sources refer to the 28nm process node as being particularly tricky. We have also heard that companies that were expecting to release ARM-based SoC chips on a 28nm process node this year have put a "first half 2012" date on that.
Perhaps AMD believes its CPU and GPU combination APU chips are more suited to tablets where the power restrictions can be relaxed. The problem here is that once again ARM vendors such as Nvidia, Qualcomm and Texas Instruments already have the lion's share of that market and even Intel, not to mention AMD, will face an uphill battle to gain traction.
AMD might have been taking heat for not going into the smartphone market, but its decision not to step into the ring with the biggest chip vendors is prudent. There is a big question that even a company of Intel's size can break into the smartphone and tablet market, and AMD might have decided wisely that its money will be better spent on parts it believes will have a better chance of selling. µ