Who will be the Intel of hand helds
Competition abounds as Intel aims to be as dominant in wireless as it is in PCs
By Heather Clancy, CRN
2:32 PM EST Fri., Apr. 06, 2001
Intel,the company that made "Intel Inside" virtually synonymous with PCs,hasn't exactly been known as an underdog in the IT arena. But when it comes to the burgeoning handheld computing market, all bets are off.
The chip giant is finding itself up against high-tech titans such as Texas Instruments and Motorola, as well as intrepid silicon start-ups, as it jockeys for the lead in setting the chip architectures for next-generation wireless and Web access devices.
Intel has huge aspirations for this product segment, but the Santa Clara, Calif., company now faces some formidable competition. Case in point: Texas Instruments supplies the silicon underlying approximately 60 percent of the cellular handset market, including products from Nokia, Sony, Ericsson and Motorola, according to market-research estimates. "That's the team to beat," says Will Strauss, principal analyst at research firm Forward Concepts.
What's more, the hardware being built around the new wave of chips for PDAs and data-enabled cell phones bears little resemblance to the desktop architecture where Intel made its name. The form factor is different for almost every device, battery life and heat dissipation are critical issues, and integrated wireless communications is de rigueur, industry experts say.
"As an implementer of 'all-channel' access solutions that encompass browser-, broadband- and wireless-based technologies, it is not so much the hardware architecture of end-user devices that is our main concern, but rather the specific characteristics and capabilities that each possess," says Edward Pearson, partner and chief architect at Quidnunc, a New York-based e-business services firm.
Handheld devices, in fact, are expanding the computing paradigm, says David Feldman, founder and CEO of Palo Alto, Calif.-based ZF Micro Devices, one of dozens of small companies hoping to provide chips for handhelds and all types of embedded systems.
"When you say 'computer,' most people think laptop or desktop. That is so far from true," Feldman says. "The 'computer' is ubiquitous. It's not [what] they describe."
The debate over system design centers on handheld computers such as the Palm series and Compaq Computer's hot-selling iPaq Pocket PC, as well as smart phones and smart-card readers. And there's plenty at stake. By the end of 2003, the U.S. market will have an estimated 107 million Web-enabled cellular phones and 8 million Web-enabled PDAs, according to research firm Gartner. In business terms, smart handhelds will form a $26 billion market by 2004, research firm IDC projects.
"The whole wireless industry is banking on making this data infrastructure happen," says Ron Smith, senior vice president and general manager of the Wireless Communications and Computing Group at Intel. "A lot of this business, quite honestly, is happening outside of the United States."
Intel is busy recruiting VARs and integrators, including the likes of Razorfish, to build business solutions based on emerging wireless technologies, Smith says. "A lot of the applications, especially, will be defined around different business segments," he says. "For any of those vertical markets, you need a solution provider to work with."
Sometimes a particular application may warrant a specific device, Quidnunc's Pearson says. For example, if the user needs to access and view large amounts of text, a handheld device might be appropriate. But if the user only needs a brief message or information alerts, a data-enabled cell phone might be more applicable, he says.
Quidnunc also is focusing on building intelligence into the server side of wireless solutions, which would enable content to be mapped automatically to various handheld devices, Pearson says. For instance, a user might see the full functionality of an auction site on a desktop PC but only the bidding function would be available via a cell phone.
"The mistake made too often is to try to deploy functions to devices that are not capable of delivering them in a usable manner," Pearson says.
Despite such cautionary tales, Intel aims to parlay its reputation in the desktop PC design world into the mobile computing field. And solution providers tackling the wireless market say the chip maker is well-positioned to do so, considering its expertise in standardization issues, cash reserves and manufacturing connections.
"They have a lot of experience and a lot of value to add to the new emerging platforms of the next decade," says Dave Hall, senior vice president and CTO at CompuCom, a Dallas-based digital infrastructure services firm.
Hall cites Intel's role in helping to solve some of the challenges that plagued early desktop PC technology, such as hardware standardization, remote management and asset management. "It's a smart time for them to be getting into this," he says.
Intel executives are thinking the same thing. They say the company's wireless chip architecture,the Intel Personal Internet Client Architecture (PCA),provides the computing power inside handheld devices, and its flash memory and communications chipsets round out an integrated solution.
The next generation of Intel's StrongARM handheld chip (part of the PCA plan) will rely on XScale technology, designed to scale a device's power consumption up or down according to the application being used on a handheld device. Compaq already plans to support the technology in future products, and Intel recently landed an endorsement from Sonera, a leading mobile-data services provider based in Finland. Under the pact, Intel and Sonera plan to pool their investment funds to back companies working on mobile applications and solutions. Intel has dedicated at least $500 million to its communications fund, which specifically backs companies working with PCA or XScale.
Late last month, Intel also disclosed plans to work more closely with IBM, which will supply an embedded version of its WebSphere middleware for Intel's PCA. Solution providers say such developments are encouraging.
"What we're all about is really leveraging current technologies," says Dana Cambra, vice president of product development at iScribe, a Redwood City, Calif.-based mobile solution provider. "When you look at the money that is flowing into the area, right now the Pocket PC,because it is more open and standard,has more extensions to it from a wireless point of view."
Still, Forward Concepts' Strauss says Intel is fighting an uphill battle as it tries to win respect in the wireless industry. Its approach to standards, for one, is radically different, he says.
"Intel is used to becoming a de facto standard," Strauss says. "They are now in a market where the standards have to be in place first, because if you don't have standards, who else is going to be on the other end of the line?"
Over the next three years, the battleground for chip architectures will revolve around development of next-generation ARM silicon, Strauss predicts. Although Intel has a solid message in that arena, its rivals aren't standing still, he says.
This year, Motorola plans to segment its DragonBall line, expanding beyond its current 68K architecture into the ARM world. The 68K line will focus on power efficiency and offer a baseline approach to handheld devices, whereas the ARM-based chip will support wireless plug-ins and other media-access technologies, says Kyle Harper, manager of strategic operations for emerging markets at Motorola. The company has dubbed this architecture DigitalDNA.
If Texas Instruments is the company to beat in handsets, then Motorola might be Intel's fiercest rival in PDAs. Motorola's chip powers current generations of the Palm product series, and Motorola estimates that it has shipped more than 11 million units to date. About 75 percent of all handheld systems shipped through the end of last year used DragonBall, according to IDC estimates cited by Motorola.
One wild card is Palm's recent decision to begin using the ARM architecture in future generations of its handhelds, leaving the field wide open for Texas Instruments, Intel or Motorola. Palm hasn't disclosed which ARM vendor it will use, and company executives declined to discuss Palm's plans.
Other chip architectures are in the handheld mix as well, including new integrated silicon from several start-ups and technology from Hitachi, which holds sizable market share in Asia, says Sayeed Choudhury, marketing director for the consumer business unit at Wind River, an Alameda, Calif.-based provider of embedded systems software. Hitachi, too, is the only major supplier of cellular handsets that doesn't use ARM, Forward Concepts' Strauss says.
Despite the challenges ahead, Intel has two big advantages in building an identity as a handheld architecture supplier: deep pockets and serious manufacturing muscle, analysts and solution providers say.
"The challenge of a hardware vendor like Intel is to provide electronics you can package into a small form factor," iScribe's Cambra says. "It requires hundreds of millions of dollars of investment to do that."
