Just months ago at the 2011 Intel Developer Forum, executives with the world's largest traditional personal computer chipmaker, Intel were all boast and bravado, saying their competitors were years behind in process. Indeed, the talk about the dramatic gains in terms of power efficiency and clock speed using Intel's proprietary 22 nm FinFET 3D-transistor design sounded very impressive.
But the first chinks in the armor perhaps began to show at the 2012 Consumer Electronics Show, when Intel was caught faking its 22 nm Ivy Bridge DirectX 11 demo during its ultrabook pitch. Intel brushed off the trickery, but the incident raised some serious questions. If the 22 nm chip was launching in April at production volume and had already been taped out in final form, why would Intel have to use canned video? Why couldn't it show its real product? Why the obfuscation??
Well, DigiTimes is reporting that multiple OEM sources have shared that Ivy Bridge is being delayed from April to June. While not a huge delay, the report raises questions about whether Intel's 22 nm process is as stable as it claims.
To be fair, the OEMs appear to be claiming that the delay is due to inventories: Because most first-tier notebook vendors are having trouble digesting their Sandy Bridge notebook inventories due to the weak global economy, while Intel is also troubled by its Sandy Bridge processor inventory, the CPU giant plans to delay mass shipments of the new processors to minimize the impact, the sources noted.
In other words, PCs didn't sell well in 2011, Intel built up a surplus of CPUs, and so it wants to delay its release. This is all very plausible, and indeed lines up with write-offs found in Intel's earnings reports.
But it is also possible that Intel isn't being entirely forthcoming and that Ivy Bridge wasn't being delivered at the reliable high volumes it had hoped. And it could very well be a bit of both factors -- too high inventories, and some struggles on the process front.
Regardless, it sounds like customers will have to wait on Ivy Bridge, a bit.
That's good news for the competition. AMD hopes to aggressively roll out its Trinity accelerated processing units (APUs) later this year. The chips are built on a 32 nm process (GlobalFoundries), but still aim to be competitive with Ivy Bridge in terms of power consumption and graphics performance. AMD is gambling that the CPU will lose, processing speed-wise, to Ivy Bridge, but be "good enough" for most consumers.
AMD hopes to price its chip + chipset package at hundreds of dollars beneath Intel. Where as Intel is targeting systems $700 and up, AMD has stated to us that Trinity systems will retail for $500 or less. Strong 2011 APU sales of AMD's initial swing at this strategy made it look like a home run.
Likewise, ARM CPU makers, including Qualcomm, Inc. (QCOM) are looking to invade laptops and compact desktops late this year, with the introduction of 28 nm ARM CPUs compatible with Microsoft Corp.'s (MSFT) new Windows 8. The Q4 2012 devices are expected to follow a strategy similar to AMD's -- strong power efficiency at a low price.
The delay is also good news for third party USB 3.0 chipmakers like Renesas Electronics Corp. (TYO:6723), ASMedia Technology Inc., and Etron Ltd. As Ivy Bridge was the first Intel chip to include on-die USB 3.0 support, it was expected to render these competitors' designs obsolete. But now, they have been bought a bit more time.
Intel's core hope in terms of maintaining its dominant position is to beat the competition in process, and trickle down its process improvements into its budget models, mitigating cost and architectural disadvantages. Intel has made big promises regarding Atom-powered smartphones, but without 22 nm technology it appears to be forgoing any sort of big mobile push in 2012. The longer it waits, the more advantage it gives to the hungry rivals. Intel should hope that the delay does not set back its very aggressive 22 nm Atom rollout.