| Last Modified On : | August 22, 2008 11:02 AM PDT |
Rate |
|
by Geoff Koch
It’s a profound understatement to say that Intel is hot on mobile computing. When it comes to on-the-go devices, the company touts its innovations in areas as diverse as people-friendly design and energy efficiency. Oh yeah, Intel is making noise about helping to weave the wireless Web that connects all these devices, too.
A tech goliath with annual revenues in excess of $35 billion, Intel is probably one of the few companies capable of backing up most of its mobility-related marketing claims. Yet despite decades of microchip dominance, even Intel does not have unlimited resources and is still subject to the occasional strategic misstep.
Case in point: for years the company worked to build a viable business around its XScale processors, low power chips optimized for rich services over wireless networks. For years, Intel slowly collected design wins in devices such as the RIM Blackberry handheld, the Dell Axim family of Pocket PCs and several handheld devices from Palm and other vendors. However in June, Intel announced it was selling its XScale business to Marvell Technology Group to focus resources on its core x86 and server businesses.
“The sale will remove a business that struggled to take off and has weighed down the company’s overall financial performance, despite billions of dollars in investments,” wrote* Dawn Kawamoto and Tom Krazit in an June 27 article published on ZDNet News.
“At one point, Intel envisioned the communication market as a hedge against the expected saturation of the PC market, Kawamoto and Krazit continued. “But building chips for PCs is far different than building chips for communication networks.”
Here it’s important to point out that mobile computing and communication at Intel encompasses far more than XScale. It’s a mistake to interpret the Marvell announcement as anything more than a clear-eyed business decision to focus on core competencies and spend resources most efficiently. The fact is Intel continues to churn out mobile- and wireless-related pronouncements, nearly all of which seem to hold up tolerably well to scrutiny and are consistent with longstanding mobility goals at the company, such as reducing power consumption.
An example: in 2006, Intel offered a five-watt, ultra-low-voltage microprocessor for use in ultra-mobile PCs, or UMPCs. By 2008, the company will be shipping chips that consume ten times less power and are one-seventh the size of today’s processors.
“So the idea of getting essentially all-day battery life out of a highly-portable, highly-functional, highly-featured devices is now very much upon us,” said Intel CEO Paul Otellini at the Fall 2006 Intel Developer Forum (IDF) in San Francisco.
What other Intel ideas about mobile computing are upon us today? And how do such ideas fit into what’s happening in the greater mobile marketplace? What follows is one attempt to answer these questions.
Though best known for it manufacturing prowess, I ntel has long been engaged in a quiet effort to influence device design. Now this work is being extended to the mobile market, as well.
At Fall 2006 IDF, Dadi Perlmutter, senior vice president and general manager of the Mobility Group, briefly demonstrated a laptop built with a strong nod to ergonomics. The device’s screen can be raised several inches above the keyboard to ease backs and necks made sore by hunching over typical clamshell devices.
There are several examples from recent years of Intel spurring new ways of thinking about seemingly tried and true technologies, often in on-stage demos at conferences such as IDF. Casual Intel observers might think such demos do little beyond generating a few predictable and soon-forgotten stories in the trade press. But a closer look shows the influence of public imagining.
Remember the ISA slots and MIDI, PS/2 and other serial port connections common to PCs in the 1990s? A 1998 concept PC loaded with USB buses helped to show* manufacturers there was a better way.
At Comdex in 2001, Intel demonstrated* a PC loaded with connectivity options ranging from an onboard LAN controller to Bluetooth to a built-in 802.11b card – all features that today are fairly common in new PCs and laptops.
So perhaps it’s worth it to pay attention to the fact that by 2005, Intel was showing off various UMPC docking station scenarios, including one associated with rear seat entertainment in cars. Already today, at least a few aftermarket UMPC car cradles exist. Based on a compelling demonstration* from Fall 2006 IDF, it’s time to start watching for cars that will ship with UMPCs pre-installed.
The point is that innovation is about more than coming up with new ways to string together integrated circuits on silicon chips or hack together novel bits of code. Especially in the mobile space, where devices and services are intensely personal, design considerations loom large.
Developers should rally behind the efforts of Intel and other manufacturers to produce mobile devices with the right combination of beauty, color and proportion; convenience or comfort in use; durability; and expression of function in terms of form. Why? Because new markets await those who get the formula right.
It’s a somewhat overused example, but remember that Apple was hardly first to the market for digital music players with the iPod. The company was just first to get all the design and ease-of-use elements right.
What’s the next design horizon? It’s impossible to predict, but consider that several companies are vying to find the right simplicity-focused solution that finally frees up digital music and infotainment generally in automobiles, according to an analysis by Julia Cosman, an Intel ISV innovation accelerator based in Germany.
“For improving the music experience in the car, there is a trend of simplifying connectivity in devices such as MP3 players, phones and portable navigation devices,” she wrote in quarterly mobility segment report circulated internally at Intel. “Quite a few manufacturers are considering offering Web 2.0-based personalized music services on the go, which work with referrals and recommendations and make a complete personal radio station for each user.”
No mobile computing or communication experience can be complete while questions remain about battery life, an issue to which Intel has devoted significant resources in recent years. Every watt counts since cutting one watt generally buys another 25 to 30 minutes in usage time.
Intel has invested considerable engineering know-how to cut power consumption from its CPUs and chipsets. Now the company is working to do the same thing to cut power across the entire platform.
One example: Intel is working with Toshiba Matsushita on display technology that automatically switches to the lower power interlace mode when the laptop is unplugged. The result is a potential savings of some 400 milliwatts.
“It’s a lot of work, milliwatt by milliwatt, but we are getting there,” said Perlmutter of Intel’s ongoing power reduction efforts in his recent IDF speech.
Better battery life is crucial to the so-called prosumer (professional consumer), who increasingly needs to be able to work anywhere. This means accessing, editing and re-uploading PowerPoint, Excel and Word files, tasks which are difficult given the present state of mobile technology.
“The only business applications properly available on mobile devices today are e-mail, voice-over-IP (VOIP) and some personal information management (PIM) applications,” wrote Cosman. “Laptops are the only mobile devices today used for office applications and to access the company intranet; people want to be able to work on the go with smaller devices.”
UMPCs, with bright five- to seven-inch displays and support for a variety of rich media and location-based services such as GPS, promise a remedy to the e-mail-only scourge afflicting most handset-type devices. Yet despite the fact that Intel has thrown much support behind UMPC development, questions persist about whether the devices will ever really be able to fulfill their promise of offering full PC functionality.
“UMPCs were designed to fulfill every role offered by every other portable device; they do everything poorly and nothing well,” wrote user conradj21 in a reply* to a question I posted on Yahoo! Answers while reporting an earlier article for the Intel Software Network. “I don't like them, I don't think they will ever be mainstream and they are just too expensive for the features they offer.”
This is reminiscent of complaints levied against many new technologies. None other than Gordon Moore – of Moore’s Law fame – once famously mused that he was not sure what PCs might be good for beyond helping housewives to store recipes.
Of course, technologies mature and customers invariably figure out new ways to process, store and share information. Cosman believes that UMPCs adoption is likely to be sped along by several imminent ultramobile usage scenarios. In the car, she said, UMPCs will be pressed into service providing everything from 3-D navigation with overlay maps and video information for top points of interest for those in the front seat to games and mobile TV or other video playback for those in the backseat.
Lest there be any doubt about Intel’s laser-like focus on mobile platform innovation, consider an abbreviated laundry list of improvements coming to Santa Rosa, the platform that will follow Centrino Duo: increased bus speed from 667 megahertz to 800 megahertz; allowing for increased throughput to memory, reduced latency and higher CPU performance; enhanced dynamic scaling of foreign performance into the bus, which effectively ratchets CPU frequency up and down to reduce active power and improve battery life; a next-generation chipset, Crestline, that will deliver richness, realism and lifelike effects in high definition playback; and NAND Flash-based disk cache that boosts application performance and execution.
Collectively, these innovations are about alleviating the many small headaches still associated with personal computing and communication. Intel’s forthcoming Robson technology, for example, allows users to save time by booting applications from NAND flash memory instead of from the hard drive. At a fall 2005 demonstration in Taipei, a laptop with Robson technology opened Adobe Reader in 0.4 seconds, while a second laptop without Robson required 5.4 seconds. The Robson-equipped laptop opened Quicken in 2.9 seconds compared to 8 seconds with the non-Robson laptop.
Intel is also working to increase the broadband wireless connections to these full-featured mobile devices. The goal is to go beyond Wi-Fi for reasons best explained by this paraphrased quip from Intel executive Sean Maloney: Wi-Fi is like pepperoni on pizza. It adds a wonderful taste but doesn’t cover the whole thing, which is why you need cheese to cover the other spaces.
The cheese, of course, is a connectivity option such as third-generation (3G) wireless, WiMAX or both. Thanks to a partnership with Nokia, the Santa Rosa platform will include a wireless HSDPA wideband CDMA card. This should accelerate the adoption of wireless WANs while also increasing the 3G attach rate among Intel-based mobile devices. And Intel continues to invest heavily in WiMAX, a technology it sees as the best solution for global high-speed wireless access.
“Our goal is to really have millions of users with their notebooks or UMPCs with our WiMAX solutions able to connect anytime and anyplace they’re going to be around the world,” said Perlmutter, describing the 2006 release of the Intel Centrino mobile technology reference guide for WiMAX.
This cresting wave of connectivity options coupled with open Web-based standards are what ultimately portend the most promise for mobile-focused developers, particularly in the area of location-based services. Already, Cosman observed, there is a whole slew of Google mashups and other Web 2.0 services that allow us ers to get personal recommendations for restaurants, bars, hotels and so on.
Another location-based application is Wi-Fi triangulation. The best example of this may be Loki*, a location-based Internet search, content navigation and location-sharing browser toolbar offered by Skyhook Wireless. The application uses Wi-Fi access points to finally allow internet users to tailor their view of the digital world based on their physical location. Skyhook Wireless has received support from Intel Capital, another example of Intel subtly helping to fuel mobile innovation.
So in the final analysis, the future does indeed appear bright for another industry inflection point, dubbed by Intel CEO Otellini as broadband-to-go. What exactly is required to meet this new market potential?
Answering that at Fall 2006 IDF, Otellini said “we need pervasive wireless broadband, and we need a new class of ultra-mobile computers that we can literally carry in our hand.”
I’m betting we’ll encounter both just around the next bend in the road.
Geoff Koch, a former Intel writer and editor, is a freelance writer, editor and Big 10 sports buff in Lansing, Mich. He recently co-authored "Open Source Development: A Platform Perspective" for the Intel® Software Network. He also wrote the "Software Developer FAQ: Intel® Virtualization Technology," which describes a few of Intel’s own virtualization efforts, and " Computer Security: The Hardware/Software Solution," based on an interview with Intel security researcher Victoria Stavridou-Coleman. Write to him at koch.geoff@gmail.com.
