Apple Buys Chip Designer

by Erika Brown, Elizabeth Corcoran and Brian Caulfield 04.23.08, 12:15 AM ET
BURLINGAME, CALIF. -

Late Tuesday, in response to questions from Forbes.com, an Apple spokesman said Apple has agreed to buy a boutique microprocessor design company called PA Semi. The company, which is known for its design of sophisticated, low-power chips, could spell a new future for Apple's flagship iPhone, and possibly iPod products as well.

The 150-person chip company, P.A. Semi, was founded in 2003 by Dan Dobberpuhl, who was a lead designer for the well-regarded Alpha and StrongARM microprocessors developed by Digital Equipment in the 1990s.

"Apple buys smaller technology companies from time to time, and we generally do not comment on our purposes and plans," said Apple spokesman Steve Dowling. He declined to comment on the value of the deal, which a person familiar with the deal suggested was done for $278 million in cash. Apple is due to announce its quarterly earnings Wednesday.

The decision to center the iPhone design around a chip that Apple could own marks a significant strategic choice by Apple Chief Executive Steve Jobs, and is aimed at ensuring Apple can continue to differentiate its flagship phone as a raft of competitors flood the market. According to a source affiliated with the chip company, Jobs and Senior Vice President Tony Fadell led the tiny group of executives who spearheaded the acquisition, which included negotiations that took place in Jobs' home.

Apple's choice is a blow for chip maker Intel, which has been trying to convince Cupertino, Calif.-based Apple to rely on Intel's chips--particularly its latest low-power line up, called Atom.

Apple has been rightfully proud of the iPhone, and has predicted that it will sell 10 million of the devices by the end of 2008. But that success has had a cost, too: Virtually every mobile-phone maker is scrambling to develop an iPhone-like device. Jobs has long asserted that Apple's greatest strengths lay in its software and in its ability to integrate hardware and software. The result: Machines that combine appealing design with an intuitive user interface, such as the iPhone. But interfaces--as Jobs well knows--can be mirrored, if not copied.

Few in the high-tech world are as wary as Jobs of turning control of core components over to a partner. The PC industry has been his proving ground; over the past three decades, he has watched numerous PC makers that have built their products around Intel's microprocessors wind up in fierce battles for narrower and narrower profit margins.

Led by Intel Chief Executive Paul Otellini, Intel is developing a line of chips that it believes can become as central to handheld computing devices as its "x86" chips have been to the personal computer industry. Intel has said it aims to create somewhat different lines of Atom processors tailored to different classes of applications, such as consumer devices, low-cost notebook computers and set-top boxes. The first of these designs is slated to begin shipping by the middle of this year; another by year's end.

But that doesn't mean Intel wants to create a unique chip for every customer--even for a charismatic customer like Apple.

Apple first got to know the designers at P.A. Semi about three years ago, when the computer maker still used PowerPC chips in its Macintosh computers. Dobberpuhl and his team, which includes engineers who had a hand in designing powerful chips, including Intel's Itanium, Advanced Micro Devices' Operton and Sun Microsystems' UltraSparc. Dobberpuhl was also the lead designer on a DEC chip project called StrongARM, which was ultimately folded into Intel after DEC collapsed. Intel tried to use the design as the basis of chips for the smart-phone industry, but eventually sold that group to Marvell.

Dobberpuhl wanted to design an enormously powerful chip, based on the PowerPC architecture, that used little power. When Apple decided to quit using the PowerPC chip in favor of Intel microprocessors, the conversations with P.A. Semi petered out.

In February 2007, P.A. Semi debuted a 64-bit dual core microprocessor which the company asserted was 300% more efficient than any comparable chips. It consumes only 5 to 13 watts running at 2 gigahertz. Telecommunications, networking and wireless companies embraced P.A. Semi's work.

Meanwhile, Apple found a collection of other chip partners to build the guts of its iPhone. One key: According to technology market research firm iSuppli, South Korean electronics giant Samsung supplies the iPhone's applications processor. Like a number of companies cranking out processors for mobile phones, such as Texas Instruments, Samsung's processor is based on a design licensed from Cambridge, U.K.-based ARM. Apple was among the three companies, also including Acorn Computers and VLSI Technology, that formed ARM in 1990.

Texas Instruments, Qualcomm and Broadcom are also fighting for share in the low-power applications processors for smart phones and mobile gizmos based on ARM designs, according to iSuppli analyst Tina Teng, with Texas Instruments leading the pack. These companies are building increasingly sophisticated processors, however, targeted at bigger and more powerful devices. "Companies like Qualcomm are trying to diversify themselves and move away from being a pure wireless company into these ultra mobile PCs, or ultra mobile Internet devices," Teng says.

Trevor Yancey, vice president of technology at IC Insights of Scottsdale, Ariz., sees ARM-based designs giving Intel a tough fight as Intel tries to push beyond notebooks and into smaller and smaller devices. "That's going to be a tough battle. If you look at the ARM core--where it has been and where it has come from--the ARM core is very well entrenched in those kinds of applications," Yancey said. "I think the gap between the two is certainly closing, but ARM is still developing at a fast pace."

Although no current Apple products use P.A. Semi chips, Apple executives kept a close eye on the work of the start-up. Talks of acquiring P.A. Semi began only in the past few weeks. Employees have been notified of the deal.

It will likely take at least a year before products incorporating P.A. Semi designs are ready. The company's flagship chip is at the heart of heavy-duty computing systems, such as those sold by NEC's storage division. But the design philosophy--hefty processing power and frugal power use--are in concert with Apple's directions. Although Apple plans to continue supporting P.A. Semi's current customers, insiders suggest that Jobs plans to use future P.A. Semi chips exclusively within Apple products.

At that point, executives believe the company will have created a unique asset--a powerful microprocessor that sips power lightly and so can support just about any imaginable applications Apple's software gurus can imagine.

Sphere: Related Content

Necessities We Can Do Without

Burlingame, Calif. -Charles Darwin would have had a field day with today's technology.

When he touched down in the Galapagos Islands in September 1835, Darwin scooped up a menagerie of finches, iguanas, rocks, plants and such. His data gathering was quick and dirty: He spent only nine days stuffing his crates with Galapagos samples and neglected to label them precisely. But he wound up with a powerful idea: Any species could mutate and evolve in response to their environmental conditions. Within the past few decades, other scientists have painstakingly documented how changing levels of food, water and other elements could cause finches to mutate in just a generation or two.

Technology, by contrast, mutates even more quickly, and we're adapting right along with it. The things we couldn't live or work without only a few years ago are now only valuable for their scrap metal. And get ready for more change. Think your laptop is essential? Guess again. Chances are that before long, you'll be tempted to give up that five-pounder for a nifty ultra-mini PC that slides into a satchel or your purse

In Pictures: 12 Necessities We Can Do Without

The managers of Xerox's famed Xerox Palo Alto Research Center back in the 1970s were among the first to articulate what brought about spurts of innovation and change in technology: Pick the biggest "resource constraint"--or the most expensive component in a system--and imagine what would happen if that became virtually free. Once upon a time, transistors cost at least a few pennies apiece. Now Intel stuffs close to a trillion transistors onto a chip that sells for only a few hundred dollars. Remember when long-distance telephone calls were expensive? Voice-over-Internet Protocol has made those chats free too.

Those kinds of innovations have meant that the gear once considered essential can be scrapped. Ready to say goodbye to your old desk phone? Even your answering machine, once an absolutely essential element of modern life, is giving way to services that companies administer in the computing "cloud"--meaning that companies are willing to store your information in enormous computer farms that you'll never see and serve it up just when you need it.

Instead of paying for a word processing system, Google Docs will let you write files and save them into the great Google cloud, where you can collect them later or send them to others. Google has even added software that lets you synchronize your Google Docs on your laptop so that you can use them in places where a network doesn't yet reach--say, on a plane.

Semiconductors continue to stuff more electronics--and thus, more processing power--onto their chips. That means that the overall devices themselves can shrink. Apple's latest generations of iPods, for instance, can be even slimmer than the original because the company no longer needs to use even a tiny disk drive but can instead stuff all those digital songs and pictures into semiconductor chips.

Novel software applications can also help cut down on the number of devices that you need, or at least, that you need to manage. Many people who carry around a cellphone haven't quite given up their office phone or even their home phone. That means technology hasn't simplified their lives; it's just multiplied the number of machines they need to answer or check for messages. Here's where software can come in: The founders of GrandCentral, a Freemont, Calif.-based start-up that was acquired by Google in 2007, created a service that redirects calls from multiple phones to a single number. No matter which of your many numbers someone calls, the calls are redirected to one number. That means you only have to worry about one phone number--and one voice mail.

Other companies are trying to apply that kind of shrink mentality to software too. For instance, to find the location of a local coffee shop, you can go to your PC, pull up your Web browser, and in a half-dozen clicks or so on a search engine, you can find a nearby coffee shop. (It helps, of course, to know your own location.)

Mike McCue, who runs Mountain View, Calif.-based TellMe, now a division of Microsoft, thinks that's too many clicks. Instead, TellMe has been developing voice recognition software that uses global positioning technology to find locations. Here's how it works: Call TellMe and ask for a Starbucks. The service starts by figuring out where you are. Then it searches Web directories for what you want. Finally, it gives you a list and a map of the nearest Starbucks. No clicks involved.

From McCue's point of view, devices should have just one button--and no stacks of menus that you have page through. "You say what you want, and we'll get the information you need," McCue says.

And that's how you really reduce the number of gadgets that you once thought that you needed.

In Pictures: 12 Necessities We Can Do Without

Sphere: Related Content

Yahoo!'s Striptease

BURLINGAME, CALIF. -
Will it be Google or AOL? Yahoo! executives are willing to team up with competitors to craft a new future for their company--provided those competitors don't include Microsoft. But Microsoft CEO Steve Ballmer still has a card to play: He's enlisted the support of Rupert Murdoch to create a partnership that would involve combining Yahoo!, Microsoft's MSN and News Corp. social-networking site MySpace.

The news marks a surprising twist in the 2-month-old acquisition saga. What's clear: Yahoo! will soon be a different company. But precisely which industry leaders will become its partners is still very much up for debate.

On Wednesday afternoon, Yahoo! issued a terse announcement that it was launching a two-week "trial advertising" partnership with Google. The arrangement is a clever way of testing two separate ideas: the financial implications of letting Google's powerful ad-serving engine take over Yahoo!'s search ads, and just how politically explosive a deal between Yahoo! and Google might be. (See: "Yahoo! Gets In Bed With Google")

That news was followed by reports that Yahoo! may also be close to striking an arrangement that would combine the search engine company with AOL, the struggling Internet division of Time Warner.

Industry pundits believe Yahoo!'s latest moves will help preserve--and even increase--its perceived share value in the face of bluster from Microsoft that it could lower its bid from the $31 per share it offered for the company Feb. 1.

In after-hours trading, shares in Yahoo! and Microsoft moved a few pennies: Yahoo! edged down by 3 cents to $27.77; Microsoft fell 18 cents to $28.71. Google rose $1.81 to $464.19.

Since Microsoft made its bid, Yahoo! executives have been considering a dizzying array of relationships with a slew of potential alternative partners to avoid a takeover: News Corp., Time Warner, AOL, Google.

Ballmer hasn't been twiddling his thumbs, either. He has reached out to News Corp., according to industry sources familiar with the companies.

Although Yahoo! has staunchly asserted it is worth more than the $31 a share cash-and-stock offer made by Microsoft, Ballmer has been reluctant to sweeten the deal. Paying more than $40 billion for Yahoo! would already send Microsoft to the credit markets to borrow money. Crafting a deal with News Corp. could lighten that financial burden as well as add the nice touch of a vibrant social-networking site, namely MySpace.

In February, Murdoch asserted he wasn't interested in acquiring Yahoo! But he didn't rule out partnering with Microsoft.

Spokespeople for the companies declined to comment.

None of this was quite what Ballmer expected would unfold when Microsoft launched its bid for Yahoo! on Feb. 1. Ballmer had hoped to quickly lock down a deal with a company that seemed adrift. However, he didn't seem to factor in the fierce antagonism that Microsoft still sparks in Silicon Valley, particularly among industry veterans including Yang and Google's chief executive, Eric Schmidt.

Microsoft executives consulted with former Hewlett-Packard executives who had led the integration of HP and Compaq for advice on how to orchestrate a clean deal. Start talking right away with managers, counseled the HP veterans. The countdown to carrying out a smooth merger starts from the time the deal is announced--not just from the time it is formally approved.

But Yahoo! executives had little interest in taking part in any kind of pre-merger discussions with Microsoft. Instead, they launched wide-ranging conversations with possible white knights who could keep the company out of Microsoft's grasp. Yahoo! also adopted measures aimed at making a hostile takeover expensive, including promising executives two years' severance if they left Yahoo! "with good reason" within two years following an acquisition.

Microsoft and Yahoo! managers have met at least twice since February. Ballmer has described the talks as "limited" and "not substantial." Yahoo!'s Jerry Yang and Chairman Roy Bostock, in a retort Monday, said the conversations were "constructive" on topics "including integration and regulatory issues." (See: "Yahoo! To Microsoft: Go Away") But maybe Yahoo! executives were simply listening for clues about how to skirt antitrust hurdles if they chose to work with Google instead of the Redmond software giant.

The trial advertising relationship announced Wednesday signaled that Yahoo and Google could be close to finalizing such a relationship.

Here's how it is likely to work: Over the next two weeks, Web surfers based in the U.S. who plug a search term into Yahoo!'s search engine will, as they always do, see a series of ads running alongside those search results. The overwhelming majority of those ads will still be served up by Yahoo!'s in-house system. But "no more than 3%" of the advertisements will be generated by Google's AdSense for Search service.

What the two companies will learn from the test is how to set a value for either an allowance that Google might promise to Yahoo! or how to structure a revenue-sharing deal.

In either case, the test will involve swapping far more information about sensitive advertising mechanics and revenues than Yahoo! has been willing to share with Microsoft.

As recently as October, Yahoo!'s Yang was reluctant to even consider turning over Yahoo!'s ad systems to Google. In a conference call last October with analysts, Yang said: "We believe having a principal position in both search and display advertising is critical to creating ... long-term shareholder value." Last year, Yahoo! generated 87% of its total revenues from "marketing services," the combination of so-called "banner" advertisements as well as search advertising

In its annual report, Yahoo! lists Google as first among its competitors: "Google’s Internet search service directly competes with us for Affiliate and advertiser arrangements, both of which are key to our business and operating results." The report also lists Microsoft and AOL, as well as other publishers, as competitors.

Antitrust experts on Wednesday were quick to note that there are no regulatory problems with the carefully limited test outlined by Yahoo! and Google. "If Google controls Yahoo in some way, it would lessen competition in the market and that's what the laws are concerned about," notes Lande, professor of anti-trust law at University of Baltimore Law School. "But of course, that's not what they've announced."

If Yahoo! chose to let Google handle a bigger chunk of advertisements, regulators might take a closer look. "A search engine is searching and advertising, and if you merge advertising, that's very much like a merger," Lande suggests. "A joint venture is evaluated under the same anti-trust laws as a merger."

"Should there be moves to make this agreement permanent, we will examine it closely in the Antitrust Subcommittee to ensure that it does not harm competition," said U.S. Senator Herb Kohl (D-WI), chairman of the Senate Judiciary Committee's Subcommittee on Antitrust, Competition Policy and Consumer Rights in a statement. "Following closely on the heels of Google's acquisition of DoubleClick, this Google-Yahoo alliance would represent even further consolidation in the internet advertising market," Kohl said.

Large Yahoo! shareholders are clearly weighing the price of the deal and how much it would help or hurt mixed portfolios that include both Microsoft and Yahoo! holdings. (See: "Ballmer's Secret Weapon")

David Hilal, associate director of research at Friedman, Billings, Ramsey & Co., told Forbes.com that he was betting that Microsoft might still be tempted to raise its bid. That move could appease a few Yahoo! executives and some big institutional investors.

Wendy Tananka and Andy Greenberg contributed to this report.

See Also:

No More Mister Nice Guy

Stop Wasting Time, Microsoft Tells Yahoo!

The Yahoo! Deal

Sphere: Related Content