On the date of publication the author of this article owned a very small amount of Intel stock in a long-term portfolio.
Twelve years ago, Intel declared war on computer cords with a campaign it dubbed “Unwire.” The $300 million ad extravaganza equated wireless connectivity with freedom. But it focused on just one wire in particular—the Ethernet cable.
Next year, Intel plans to finish the job by eliminating everything from charging cords and display cables to wires for all peripherals. “Our vision is very simple: we are going to get rid of all that,” Kirk Skaugen, senior vice president and general manager for the PC Client group, told Intel developers last month.
The key to making Intel’s vision a reality is a new chip architecture code-named Skylake. Like the current Broadwell architecture, Skylake will use a 14-nanometer manufacturing process. But it also promises major performance gains and gee-whiz features like the ability to charge with wireless power and automatically connect PCs with displays, USB storage devices and more. Engineers anxious to get started building on Skylake can get a wireless charging development kit in the next month or two. A Skylake reference design will be ready in the first quarter of next year, and volume production is expected to follow six months later.
Intel is betting the new platform will spur a new cycle of PC purchases, much like the introduction of Centrino did a decade ago. “The interesting thing about about these kinds of technologies is they impact PC refresh,” Tom Garrison, vice president of Intel’s PC Client Group said. “As the refresh rate increases that drives our business results.”
The impact can be dramatic. Within two years of its release Centrino had lifted revenues of Intel’s mobility group by 161 percent. Even as the average selling price for other chips was declining, Centrino was able to command a premium. By 2005, Centrino had contributed $2 billion in revenue, according to Intel internal reports. Intel’s stock, which opened at $16.44 on January 3, 2003, closed at $24.96 on December 30, 2005.
Will history repeat itself? To say investors are skeptical would be an understatement. Last week, Morgan Stanley downgraded Intel on concerns of an inventory buildup. In a research note, analyst Joseph Moore called the company’s unexpectedly strong third quarter and rosy fourth-quarter forecast “as good as it gets.” Intel’s stock fell more than ten percent, though it has since recovered.
There’s no denying the “No Wires” message has broad appeal. Since wireless charging first became available on the Palm Pre in 2009, it has become the perennial tech tease. Wireless charging’s arrival in mainstream consumer products has been promised so often that comedians like Stephen Colbert have turned it into a running joke. When Tesla founder Elon Musk asked Colbert earlier this year what he would most like to see invented, the comedian replied: “I wish there weren’t any cables...cables on anything is terrible.” Colbert mused that he should be able to walk into his house, and have things “just charge,” or have a “charging subscription” that followed him around the United States. “Ok, ok, we’ll do it,” Musk responded.
Old Technology Is New Again
The punch line, of course, is that it’s already been done. The science behind wireless charging was described in 1831 by Michael Faraday and demonstrated by Nikola Tesla in 1891. But transporting electricity via a wire proved more practical. It wasn’t until people grew umbilically attached to their smartphones that researchers turned to wireless charging as one way to satisfy the constant demand for energy from growing numbers of mobile devices.
Marin Soljacic, a professor of physics at MIT, became obsessed with wireless charging after being repeatedly woken up by the beeping of the dying battery on his wife’s cellphone. Together with a small team of researchers, he figured out how to tune a set of coils to efficiently transmit and receive energy by using a specific frequency. In 2007, the team demonstrated they could light up a 60-watt bulb from a distance of seven feet. Shortly afterwards, they founded a company, which they named WiTricity, to continue refining the technology.
Meanwhile, researchers around the globe were racing towards their own wireless power breakthroughs, with teams working everywhere from Auckland, New Zealand to Jerusalem, Israel, Cambridge, UK and Seattle, WA. In August of 2008, Intel’s chief technology officer, Justin Rattner repeated MIT’s light bulb demonstration at the Intel Developer Forum and revealed that Intel was working on adding the technology to laptops. The lead researcher, Josh Smith, now a professor at the University of Washington, tried to set realistic expectations. He told a reporter: “There are a lot of open questions, and a lot of really hard engineering work that would be required to turn this into a product.” But all most people heard was that wireless charging was coming computing devices.
In fact, it would take another four years for wireless power in low-power smartphones to hit the U.S. market. The first medium-power products only arrived in October when Bosch started selling power tools that charged wirelessly. Six years after Intel started working on the problem, PCs are still tethered to their charging cords. “The ability to charge lower-power devices was better understood,” Intel’s Garrison explained when I asked about the delay.
“This is a long-term initiative; it’s not going to come about tomorrow,” said Angelo Zino, an equity analyst who follows Intel for S&P Capital IQ. Zino is ok with that. When Intel does finally move into wireless charging, he said, it will be a big deal and could allow the company to see growth longer term. “Intel is looking for ways to innovate, and I think the “No Wires” initiative is a way for them to innovate on the PC side of things,” he said.
A Low-power Gamble
Between 2011 and 2013, Intel’s revenues fell two percent. It wasn’t a huge drop but in the context of a rapidly weakening PC industry it was a frightening one. “Computer Sales in Free Fall,” proclaimed a headline in the Wall Street Journal in April 2013. The article blamed a 14 percent quarterly decline in worldwide shipments of laptops and desktops on the advance of iPads and other touch-based mobile devices. Since then, the PC market has stabilized, but not before the two-year decline set the stage for Intel to revisit the launch of Centrino, widely see as one of Intel’s marketing triumphs.
Two years before Intel launched Centrino the PC market contracted for the first time ever, and Intel’s annual revenues nosedived 21 percent. A general awareness that Intel was focused on the wrong thing crystallized into a strategic change so abrupt it would later be described as “the right turn.” For years, Intel had emphasized increases in processor speeds, equating faster clock speeds with better performance. But faster chips consumed more power and generated ever-increasing amounts of heat. This created major problems for laptops where longer battery life and smaller sizes were the characteristics most valued by consumers. Intel was focused on manufacturing Lamborghinis when all most customers wanted was a Honda.
For the first time, a controversial project code-named Banias appeared on Intel’s roadmap. Banias, which had begun as an experiment by Intel engineers in Israel, was a repudiation of Intel’s clock-speed doctrine. Instead of striving for greater gigahertz, engineers working on Banias sought to balance performance and power. They turned off by default all parts of the chip that weren’t needed and squeezed more instructions into each cycle. The actual size of the chips shrank.
The experiment paid off. Pentium M, the chip at the heart of the Centrino platform, ran about half fast as Pentium 4 chips for desktops, but it performed just as well, and in some tests even better. It also offered significantly improved battery life. “This chip really cooks along, just as Intel promises,” wrote David Pogue of the New York Times. Anand Lal Shimpi, publisher of AnandTech, a site respected industry-wide for its detailed reviews and benchmark tests, declared the Pentium M to be “unbeatable” for business/office applications.
PC makers like Dell and Hewlett-Packard were likewise impressed by the Pentium M, but they were less excited about the wireless technology that Intel insisted they buy as part of the entire Centrino platform. “We weren't close to the idea that we would build every single notebook with Wi-Fi,” said Alex Gruzen, then the head of HP’s mobile division. “Up until that time Wi-Fi had been an optional feature.” It also didn’t help that Intel’s wireless technology wasn’t considered to be as good as products that were available from competitors. But perhaps the biggest issue was the lack of existing Wi-Fi infrastructure that would make the new Centrino-branded computers compelling to consumers. According to Gartner Dataquest, in 2002 there were only 14,000 public hotspots worldwide.
Inside Intel, the shortage of hotspots was referred to as the “red-face” test — the minimum amount of infrastructure that was needed in order for the company to be able to stand up and proudly proclaim that wireless connectivity with Centrino would be widely available. “In September 2002, we had a review with Craig Barrett (Intel’s CEO) and Paul Otellini (executive vice president of the Intel Architecture Group),” recalled Yoav Hochberg, who was then the director of Intel’s mobile group. “We said we could have 400 verified hotspots by Centrino’s launch. Craig and Paul said you are right, but you are missing a zero.” By March, Hochberg had signed 179 deals involving hotel chains like Marriott, restaurant chains like McDonald’s and retailers like Borders. Momentum continued to build throughout the year and into the next. The number of verified hotspots reached 30,000 by the end of 2003 and 100,000 by the end of 2004.
As hotspots spread, so did consumer enthusiasm, helped by an advertising blitz that suggested wireless computing was the element that could transform otherwise ordinary, humdrum lives. The ad campaign began in early March with TV spots that suggested a major breakthrough: “On Monday 12, Intel will not only change how you work, but where you work.” Print, outdoor and online ads followed, with exclusive placement on CNET and CBS Marketwatch and advertorials in PC Magazine and the New Yorker. Wired ran a 64-page special issue called “Unwired” that was sponsored by Intel. Road warriors were targeted at airports and train stations, where they were offered hands-on experience with Centrino-enabled laptops. On September 25, companies like Cisco and Boingo teamed up to offer free wireless access at hotspots in a worldwide festival that was dubbed “One Unwired Day.”
Needed: An Allied Force
Intel’s executives—and the company’s industry partners—have been thinking a lot about Centrino as they plan for the release of Skylake. “Centrino was probably one of the most successful things we’ve done in the PC industry, and that was really just about eliminating a single cable, the Internet cable,” Kirk Skaugen told developers in September. It was, remembered Alex Gruzen, now CEO of WiTricity, the turning point for wireless communication. Before Centrino, “lots of people were still using PCMCI cards,” he recalled. Within a few years, it became difficult to buy a notebook without wireless built in.
WiTricity and Intel signed a licensing agreement in June, and Gruzen is hoping that Intel’s No Wires campaign will give wireless charging the same boost that Centrino gave to Wi-Fi. It’s unclear, however, if Intel is ready to make the same commitment. Garrison told me that Intel is still evaluating the No Wires marketing plan. “The things that go into that decision are really around how is the industry is evolving and moving forward,” he said. “The reality is as we go talk to customers everyone, without exception, is very intrigued with the idea of wireless charging because it solves so many different challenges.”
Companies like McDonald's and Marriott, who were part of the early Unwire campaign, don’t need coaxing from Intel to roll out wireless charging. McDonald's started testing wireless charging in select restaurants in Europe last year. In October, Marriott began offering wireless charging to guests in their lobbies.
From Intel’s standpoint, this is fantastic. Nathan Brookwood, a research fellow at Insight 64 and a longtime Intel watcher, questioned how much unique advantage Intel itself will derive from wireless charging and whether a major marketing push might benefit competitors more. “Intel is not doing it by themselves, this time,” he said, noting that Intel is part of a consortium known as the Alliance for Wireless Power, or A4WP, which includes competitors and peers like Qualcomm, Samsung, Texas Instruments, Broadcom and Marvell.
The A4WP, which supports a standard known as Rezence, takes a low-key approach to marketing. Anything similar to a Centrino-style campaign would have to come from members themselves, said Geoff Gordon, chair of the marketing committee. This wouldn’t necessarily bode well for wireless charging moving into everyday products, except for the fact that a competing standard is showing steady organic adoption.
A group called the Wireless Power Consortium, which supports a standard known as Qi, currently counts a worldwide infrastructure of one million charging stations located in places like the Capital International Airport in Beijing, China and railroad stations and airports in Japan. The problem is that only Qi-enabled devices can reliably charge from Qi-enabled infrastructure. There are technical workarounds, such as multimode chips, but compatibility isn’t guaranteed. John Perzow, the vice president of market development for the consortium, said harmonizing all the competing standards, including one from a third group, the Power Matters Alliance, is more important for the future of the technology than a consumer-focused campaign at this point in time. Perzow noted that the battle over the Wi-Fi standard had already been won when Centrino was launched. “With wireless charging, we are still slugging it out,” he said.
Harmonization efforts are underway. The A4WP and Power Matters Alliance agreed to consolidate their specifications in February of last year. Everyone claims to want to work together, and from a technology standpoint, the standards are not that far apart. By the time Intel gets around to fielding its Skylake chips, the standards battle may be over. All that’s left to conquer will be consumers' hearts and minds, and Intel is unlikely to face much resistance there. This time around, instead of fighting a war on wires, Intel may show up just in time to lead the victory parade.
Follow me on Twitter for curated tech-related news and humor.
