Meanwhile, the plotters at Sunnyvale are ahead of schedule on the low end Fusion front, while insiders have confirmed to us that the big guns, like the Llano APU, are on track for a volume ramp in the middle of the year.
While it’s true that the microprocessor wars are no way as interesting as they used to be in the days when the Megahertz Wars ruled the roost, 2011 promises an interesting divergence between technology provided by AMD and Intel.
There’s no doubt that as far as technology is concerned, Intel excels. Intel will swiftly move to Sandy Bridge CPUs, which uses a 32 nanometre, and at last September’s IDF, made it clear that we can expect 22 nanometre technology to start arriving at the end of next year.
AMD was pretty bullish about its Fusion tech at an analyst day in the US last week – the first processors, intended for netbooks and for notebooks, are set to deliver some revenues in the fourth quarter of this year. There are already Fusion chips out in the field and sooner or later one of the leaky sieves that constitute AMD’s customer base is bound to spill the beans on some all important benchmarks.
AMD seems at last to be in a relatively sound financial position after divesting itself of its fabs to GlobalFoundries. Despite the prophets of doom describing its acquisition of ATI as a disaster in the making, AMD has done well on the graphics front. I was one of those prophets of doom. I was wrong. Fusion promises great battery life and that still is one of the weaknesses of notebooks. And at the same time, AMD still seems to continue to have design wins for its processors, with quite a range of its customers subscribing to the “Vision” thing.
And what of ARM? We’ve still to see a plethora of tablets and machines based on the plucky little British company’s microprocessor technology. But they’re going to come as sure as eggs is eggs, and personally I’m looking forward to the variety of gizmos we’re going to see. If Intel has an Achilles heel, it is that it’s betting on Atoms and future Atoms to power all sorts of low power devices. I’m not certain this is going to happen and this element of Intel’s strategy could well be a serious misstep.
Roll on 2011, and let the battle between Tweedledum and Tweedledee commence.