|
INTEL'S SMALL AMBITIONS -- Diminutive form factor; big dreams. The chipmeister is counting on a new type of computer for growth.
(Information Week Via Thomson Dialog NewsEdge) Like an Old Testament prophet emerging from the wilderness, David Perlmutter came out of the Intel Research Center at Haifa, Israel, last summer to save the chipmaker's, um, bacon. One of the design brains behind both the Pentium and Centrino lines of microprocessors, Perlmutter-a gruff Israeli known as "Dadi"-introduced the Core 2 Duo chips last July. Not only did the processor line launch a turnaround for Intel, winning back market share from Advanced Micro Devices and adding nearly 28% to the company's share price in less than 12 months, it represented something of a crossover dribble for Intel: Instead of cramming more and more transistors onto ever more powerful chips and thus increasing power consumption and heat output in an engineering vicious circle, Perlmutter introduced a new design paradigm, making the fastest and most powerful chips possible within a tight range of energy requirements and temperatures.
Now Perlmutter is in charge of Intel's mobility group and, along with his lieutenant, Shmuel "Mooly" Eden, he's being asked to reorient Intel again. Envisioning a computing environment populated with devices known variously as ultramobile PCs (UMPCs), mobile Internet devices (MIDs), and even "smartphones on steroids," Intel is introducing a chip architecture to enable Internet-focused, multimedia-enhanced, handheld gadgets powered by light, cool, energy-stingy chips.
It's Intel's biggest repositioning since its drive into notebook-specific processors in 2000, which led to the Centrino chip. It's a shift that's been in the making for more than two years, when Intel put a team in place to evolve its architecture to power and cost levels "beyond what we were doing on our mainstream products" for notebooks, desktops, and servers, Perlmutter says. The world's largest chipmaker, he adds, is "rethinking the high end of the smartphone category."
The new ultramobile architecture signifies more than just a strategy shift for Intel. It's a gamble on an emerging category of devices aimed at business users, one that's met with mixed success in the past (remember the EO, from AT&T?). It's a category that conflates the capabilities of wireless laptops and smartphones in more costly, less wieldy devices that-theoretically, at least-will nudge aside the wildly popular BlackBerry and its rivals. What does Intel bring to that table? First, Perlmutter's low-profile, energy-efficient chip design; second, Intel's low-cost commodity manufacturing expertise; and third, a brand name in the computer industry that can drive market acceptance like few can anymore.
If Perlmutter and Eden are right, they're forging a new era of Internet-everywhere, long-battery-life devices that will capture big corporate accounts and lead Intel's growth into the next decade. If they're wrong, well, the seaside city of Haifa is a lovely place to retire.
SHIFT HAPPENS
The first batch of devices to use the new Intel architecture includes the Shift, a mini-PC with a 7-inch tilting screen and a full slide-out keyboard, unveiled earlier this year by Taiwanese handset maker HTC. At Intel's Developer Forum in Beijing in April, HTC said the Shift will use Intel's Ultra Mobile Platform 2007. Originally code-named McCaslin, the Ultra Mobile Platform 2007 features A100 and A110 processors along with the 945GU Express chipset for notebook-like graphics capabilities and the ICH7U I/O Controller Hub. UMP 2007 will give HTC's Shift, to go on sale in Europe and the United States in the third quarter, as well as forthcoming UMP-based devices from Aigo, Fujitsu, and Samsung, 40% more battery life compared with Intel's notebook chipsets, the company says.
For years, we've been hearing how smartphones eventually will replace computers for mobile professionals, yet only 1% to 2% of corporate e-mail accounts worldwide are available on mobile devices. Still, the smartphone market is roaring ahead: Gartner says shipments will hit 100 million this year, up from around 73 million units in 2006, and they could increase by another 50% to 60% in 2008. And corporate acceptance is growing: Business smartphones, which represented 17% of the market last year, will account for 19% this year and 24% next year, Gartner says.
As for the hotly anticipated iPhone, in debuting the sleek handset in January, Steve Jobs said Apple's goal for the first year is 1% of the worldwide mobile phone market, or 10 million units. Some analysts say that's low: Bernstein Research predicts Apple will sell 7 million iPhones in the last six months of this year alone, and another 15 million next year. How many of those will be for business use is anyone's guess, although Jobs' announcement last week that the iPhone will be open only to Web-based applications from third-party developers is likely to slow business uptake.
It's worth noting that Intel has stumbled before in trying to move into the handheld realm. CEO Paul Otellini sold off the ill-conceived XScape cell phone processor division, arguably the biggest failure in the company's history, in June of last year.
For a measure of the challenge Intel faces-along with HTC, Samsung, and the handset makers rushing to market with ultramobile devices-take a look at Jeff Hawkins, the founder of Palm Computing. Onstage at the D: All Things Digital conference last month, looking as endearingly gawky as ever, Hawkins tried to sell a skeptical crowd on the merits of the Foleo, the new Treo companion with which Palm Inc. hopes to revive its flagging prospects. Hawkins handled the Foleo, essentially a slimmed-down notebook, less like an MID than an IED, as if it might blow up in his fingers at any moment. And judging from the critics' reaction to the new device, the Foleo already has exploded in Palm's face.
"The biggest new product disappointment of 2007," says Gartner analyst Todd Kort. "I ask myself if the folks at Palm spent five minutes interviewing executives about what they really want and need," adds "Mchacur," a commenter on the mobile technology blog Engadget.
Therein lies the danger of trying to carve out a new device category, particularly one aimed at business users: The salient question can easily morph from "Will mobile professionals embrace this form factor?" to "Why should they give a damn?"
Perot Systems has deployed around 2,100 BlackBerrys and other smartphones, and it's standardizing on the Windows Mobile 6.0 operating system with ActiveSynch for corporate e-mail access, says CIO Mike McClaskey. "You can do a lot of pretty neat stuff on Windows Mobile 6 and on BlackBerry, plus still have connectivity to your corporate data," he says.
As for the emerging ultramobile category, McClaskey hears a lot of noise but no clear business case. "You start talking about these other devices, and the UMPC fits in the middle someplace," he says. "Certainly being an IT company, we have people who want the latest, greatest, coolest thing ... but I don't see us broadly deploying UMPCs at this point."
WISHFUL THINKING?
Nevertheless, hardware vendors sense a market opportunity. In the promotional material for its forthcoming Q1 model, Samsung calls attention to the limitations of conventional smartphones, PDAs, and personal media players-insufficient bandwidth, poor viewing and sound quality, and trouble syncing with PCs-which have created "a desire in the market for convergence."
A tablet-style device that runs Windows XP Tablet PC Edition, with a 1-GHz Pentium processor, the Q1 offers Ethernet, Wi-Fi, and Bluetooth connectivity, voice-over-IP capability, a 7-inch screen, and a multimedia feature that lets the user play music or videos without booting Windows-all in a device weighing less than 2 pounds. The Q1, however, doesn't have a keyboard (an attachable one is available as an option), and it carries a hefty retail price of around $1,100. You can get a very nice, very light, and very powerful Dell notebook for hundreds of dollars less-and you don't have to carry a separate keyboard.
That price disparity isn't lost on Perot's McClaskey. "The question will be the cost of [the UMPC] versus the standard laptop configurations from corporate," he says. "That's still a big premium to use one of these as a desktop, compared to what we get with Dell or Lenovo. How do you value that?"
Enter Intel. Newer subnotebook devices already cost less: The Foleo (which is being marketed as a smartphone "companion," not a standalone device) will cost $500. The new machines envisioned by Intel will be around the same price, says Rob DeLine, marketing director for Intel's ultramobile group. Along with low power, low price is a priority for Intel. "The iPhone will cost that. An unsubsidized BlackBerry costs in that area," DeLine says. "So there's a [device] budget available in that space and we've got to work within that."
Working within those price constraints, in a business already squeezed by low-cost Chinese competitors, is "fraught with peril" for the world's largest chipmaker, says Carmi Levy, an analyst for Info-Tech Research Group. Per-unit prices, both for the mobile hardware and the processors and chipsets that power them, are significantly lower than those for conventional PCs, Levy says. "This reduces margins," he says, "and makes new semiconductor platform business cases harder to justify."
What's more, product life cycles in the mobile end of the computing business are brutally short, "which only exacerbates a tenuous financial situation," Levy concludes.
ALL IN
Financially tenuous or not, it's a market that Intel's competitors are jumping into, too. Just weeks after Intel's ultramobile announcement in Beijing, AMD said at Computex in Taiwan that it will produce a new chipset, code-named Bobcat, for UMPCs. No details were offered, but Bobcat will likely be a slimmed-down version of AMD's upcoming notebook platform, known as Puma, which includes the new Griffin processor, the first CPU from AMD that's designed specifically for mobile applications.
Qualcomm, whose chips power most of the world's CDMA cell phones, is entering the market as well, with a new platform known as Snapdragon that will add longer battery life and broadband connectivity not just to subnotebook computers but consumer electronics as well. Asia's chipmakers aren't far behind: Taiwanese semiconductor company VIA Technologies said earlier this month that it will move aggressively into the UMPC market and is in talks with hardware vendors to include its chipsets in devices that could hit the market later this year.
Still, Intel claims to have the technology edge for the foreseeable future. Early next year, devices with Intel's next-generation ultramobile platform, known as Menlow, will begin to appear. Menlow comprises the new Silverthorne processor, featuring a new 45-nanometer design and new "high-K" materials to replace silicon dioxide ("K," the Greek letter kappa, refers to the ability of a material to hold an electric charge), reducing transistor leakage while increasing performance. Designed from the ground up for subnotebook devices, Silverthorne's use of high-K materials is "the biggest change in transistor technology since the introduction of polysilicon gate MOS transistors in the late 1960s," said Intel co-founder Gordon Moore earlier this year.
As the size of a transistor shrinks to 45 nanometers, explains DeLine, it gets more and more difficult for it to hold a charge. "You've got leakage, and leakage means lower battery life, which is unacceptable in this class of device," he says. Intel has "pushed through a wall" with the high-K metal gate, he says, which will support "silicon sizes and power levels that enable us to get into that [ultramobile] form factor."
Form factor, says Eden, is one of four requirements Intel designed around in the move from desktop computers to notebooks, along with increased performance, long battery life, and wireless connectivity. To those four "vectors" Eden now adds a fifth: personalization, the tendency for users to combine their business and personal lives-calendars, photos, music, video, connectivity, etc.-on one mobile device. Those devices, Eden predicts, will be Intel's "growth engine" going forward.
THE RISK
Ultimately, the issue for Intel and the other vendors pushing into the subnotebook category is the C-V-D question: convergence vs. divergence. On one hand, you can view the mobile Internet device (like the one from EB that Perlmutter is holding on p. 36) as the single converged handheld device that will combine all the power and features of a smartphone, a high-end cell phone, an MP3 player, and a laptop. On the other hand, you can see it as just another device to lug around. To the degree business users view UMPCs as the former, Intel's gamble will succeed. If it's the latter, well, Perlmutter's aware of the problem. "The risk is that this isn't so much a category that people really need, but an in-betweener that nobody needs," he says. "We're not ignoring that risk. We think this is a chance to re-create the whole smartphone category, to give it more life and more enthusiasm."
John Straub, VP of IT at Lafayette Federal Credit Union in Kensington, Md., says he's watching the ultramobile evolution, but not yet diving in. "To a large extent the BlackBerry does what we need it to do," says Straub, who manages around 20 BlackBerrys used by executives and field personnel for the credit union. "But if you had a screen that was three or four times bigger, and full Internet access wherever you are, in something that you can still carry around easily, I'd say that has a lot of potential."
Then Straub drops the name that looms large over this market category: iPhone. The forthcoming Apple device, says Straub, "is probably a more useful version than the mobile Internet devices that Intel or Samsung have in mind."
Like Straub, many IT pros are intrigued by the iPhone's combination of Apple's Safari browser plus the company's legendary multimedia capabilities. Whether the iPhone, with its sub-3G connectivity and its AT&T-only service model, will catch on in the business world remains to be seen. But the buzz over the device, expected later this month, is overshadowing awareness of other products in the category-both smartphones and smartphones on steroids.
The buzz may overwhelm Intel's strategic push, or it may prove to be what carries this product category forward. Either way, having already made its bet, Intel's in it for the long haul. "The design cycle is relatively long compared to the market, so we have to have this magic ability to predict what people would like to have three or three-and-a-half years from now," Eden says. And he points out that Intel made the same kind of bet on Wi-Fi technology four years ago: Now 96% of laptops are Wi-Fi-enabled.
To get to that kind of market acceptance with the unknown, untried ultramobile PC, Intel is going to need that magic now as much as it ever did.
Write to Richard Martin at rmartin@cmp.com.
---
informationweek.com
MORE THAN ULTRAMOBILE
Intel aims both high and low: informationweek.com/1138/intel.htm
COMPETITORS CATCH UP
Will AMD's quad core be a phenom? informationweek.com/1138/blog_amd.htm
MEDIA MAVEN
Intel unveils a multmedia chipset for Vista: informationweek.com/1141/intelchip.htm
IMAGE GALLERY
Intel's spring analyst meeting, a pictorial tour: informationweek.com/1138/gallery_intel.htm
Copyright 2007 CMP Media LLC. All rights reserved.
Copyright 2007 CMP Media LLC
[ Back To Homepage ]
|
Home
Featured Article Archives
eNews
RSS FEED
