Chip Giants Look To Tablets, Smartphones For Growth

Market research firms Gartner and IDC reported on 13 April that the worldwide PC market in the first quarter saw its first year-over-year shipment declines for the first time in six quarters, in part due to the popularity of tablets. With Apple’s iPad 2 now on the market, and a host of new tablets coming from the likes of Research In Motion, Samsung, Motorola, Dell and Hewlett-Packard either in store or on their way, tablets could continue to be a force in the PC space.

Tablets weren’t the only reason for the lower-than-expected first-quarter numbers – IDC analyst Bob O’Donnell said the extended PC lifecycles and lack of “compelling new PC experiences” also played a role – but sales of both tablets and smartphones are only projected to grow, and grow fast.

As executives with both Intel and AMD prepare to announce quarterly earnings this week – on 19 and 21 April, respectively – the health of the PC market that the two currently dominate will cast a large shadow over the companies.

Weak demand

“Weak demand for consumer PCs was the biggest inhibitor of growth,” Gartner analyst Mikako Kitagawa said in a statement. “Low prices for consumer PCs, which had long stimulated growth, no longer attracted buyers. Instead, consumers turned their attention to media tablets and other consumer electronics. With the launch of the iPad 2 in February, more consumers either switched to buying an alternative device, or simply held back from buying PCs.”

The numbers from Gartner and IDC caused concern among financial analysts, some of whom said in research notes that they had lowered their estimates for both chip makers and at least one – Romit Shah, an analyst with Nomura Equity Research – predicted that the situation could impact the first-quarter numbers from Intel and AMD and their forecasts for the second quarter.

Currently a large majority of mobile devices – including tablets and smartphones – are powered by low-power chips designed by ARM Holdings and built by the likes of Qualcomm, Texas Instruments and Samsung. Other chip makers, such as Nvidia, also are signing up with ARM. The projected numbers in the market explain why. Gartner is forecasting that tablet sales will grow from almost 70 million this year to 294 million in 2015. At the same time, market research firm In-Stat is predicting 850 million smartphone sales in 2015.

Analysts expect the PC market to keep growing – IDC’s O’Donnell expects a rebound of sorts in the second half of the year – the real movement is in mobile devices, and Intel and AMD want in.

Intel has been outspoken in its intentions – President and chief executive Paul Otellini declared in October 2010 that the company would make a major move into the tablet space – but the company’s Atom platform has been unable to meet the low-power numbers necessary to compete with ARM. Until recently, that is.

At the Intel Developer Forum in Beijing on 11 April, the company unveiled the new Atom Z670 – formerly codenamed “Oak Trail” – a 45-nanometre chip aimed at the tablet market that is 60 percent smaller than previous Atom chips, offers better battery life and comes with graphics and the memory controller integrated onto the compute die. In addition, Intel executives said they were stepping up the cadence of Atom releases, with the 32-nm “Cedar Trail” to start sampling later this year and shipping in 2012, and a 22-nm chip coming in 2013.

They also said they expect as many as 35 designs to start rolling out in May based on Oak Trail, from such OEMs as Lenovo, Fujitsu and Motion Computing, and running the Windows, Android and MeeGo operating systems.

AMD, which has been less aggressive in pursuing the tablet space, reportedly is looking to hire Android driver development engineers as it steps up its efforts in both tablets and smartphones. In the recent past, AMD executives have said they would eventually pursue a tablet strategy, but that there was no hurry. However, that reluctance to enter the mobile device space reportedly was a core disagreement between the board of directors and Dirk Meyer, who resigned as chief executive in January.

Meyer’s resignation came soon after AMD unveiled the first of its Fusion processors, which offers discrete-level graphics and the CPU on the same piece of silicon. AMD officials also have said that the company’s “Brazos” APUs (accelerated processing units), aimed at lightweight laptops and netbooks, also can power tablets.

According to analyst Jack Gold, ARM should continue to hold a commanding position it mobile device space for the rest of the year, but he expects Intel will begin gaining traction starting next year, as it ramps up its Atom innovation and moves quick from 45 nanometres to 32 and then 22. In a research note on 14 April, Gold said he expects that by the time it hits 32 and 22 nanometres, Atom will match ARM chips in energy efficiency, and noted that Intel not only will increase the core count, but include other hardware-based features around such areas as security, disaster recovery and synching.

Tablet boom

By 2015, tablets sales will be in the 200 million range, and that Apple – with its own ARM-based chips – will have a market share of 35 to 45 percent. That will leave more than 100 million units primarily running Android – with some Windows-based devices and others from RIM running QNX – and given Intel’s manufacturing prowess and aggressive roadmap, it could grab 20 to 30 percent of the market over the next three to four years, he said.

Smartphones will take longer for Intel to gain ground in – Gold said it will be another one or two generations of Atom before it can compete with ARM-designed chips. But, he said, Intel will compete.

“ARM’s competition with Intel is not intense yet, but will become so in 2012,” Gold wrote. “That’s when Intel has competitive silicon for phones to supplement its chips for tablets available this year. Intel has the manufacturing clout to make Atom chip prices attractive and potentially buy market share. We believe the marketplace, which has often been sceptical of Atom’s chance of success, is discounting Intel much too heavily at this point, given that the battle is just now getting underway and there are no clear long-term winners for next-generation technology needs.”