Despite the fact that Apple has yet to produce an iPhone chip based on its own design, and that IBM doesn't design smartphone chips, the judge overseeing the Mark Papermaster noncompete case views the two companies as chip competitors.
Judge Kenneth Karas of the U.S. District Court for the Southern District of New York filed his opinion Monday (click here for PDF) on why former IBM executive Papermaster should not be allowed to join Apple as head of the iPhone and iPod hardware engineering team. Karas' decision to grant a preliminary injunction preventing Papermaster from working at Apple was revealed a few weeks ago, but the reasoning behind the opinion was delayed until IBM and Apple had a chance to review the opinion to make sure it did not disclose any confidential information.
There's no dispute that Papermaster signed a noncompete agreement in 2006 that would forbid him from working at any company deemed a competitor of IBM's for a year following his departure from Big Blue. Papermaster's lawyers are attempting to argue that since the only overlapping product between Apple and IBM--servers--is one that wouldn't be part of his official duties at Apple, and since he won't be running Apple's P.A. Semi chip design team, the noncompete shouldn't apply.
Judge Karas appeared to agree that since Apple's server business is such a small portion of its business and Papermaster will have nothing to do with that group, that experience isn't really at issue. But in his opinion, IBM and Apple are competitors in the chip market because both companies produce or will soon produce chips that wind up in mobile phones--regardless of whether those chips are similar or even whether those chips were designed by company employees.
Of course, the court recognizes that IBM does not sell MP3 players or cell phones that compete with the iPod or iPhone. But, IBM does sell the microprocessor technology that provides the electronic brains for those products and competes for that business. To profit from the manufacture and sale of such products, IBM relies heavily on its "Power" architecture, and has employed Mr. Papermaster as its top expert in the development and application of that technology.
Karas leans on a declaration filed by Rodney Adkins, IBM's senior chip executive, in forming his opinion that IBM's chip group competes for design wins in products like the iPhone and iPod Touch. Adkins wrote, "Steven Jobs, Apple's CEO, told the press recently that 'P.A. Semi is going to do system-on-chips for iPhones and iPods.' IBM designs and manufactures microprocessors suitable for each of those applications." System-on-chip, or SoC, is a term used to describe a single chip that comes with all the technology needed to run a system, such as the applications processor, communications hardware, and other vital parts.
That led Karas to believe that IBM has a healthy business selling similar SoCs for mobile phones or iPods. "Apple announced its intention to have P.A. Semi develop the very type of product that IBM sells to the market generally, and would like to sell to companies like Apple," Karas wrote in his opinion, referring to Adkins' statements.
But IBM doesn't appear to have any customers for those mystery microprocessors referred to by Adkins as potential products for the smartphone market. An IBM representative was unable to provide the names of IBM-designed microprocessors or SoCs for smartphones or handheld computers.
And the Power architecture--where Papermaster's expertise lies--is not a serious player in smartphones or handheld mobile computers like the iPod Touch; the ARM architecture dominates this market. No major smartphone maker uses a Power-architecture applications processor in its phone, and as far as I can tell, none is really considering it.
Power-architecture processors are generally used in products other than modern smartphones or handheld computers.
(Credit: Power.org)Power.org, the industry organization dedicated to advancing the Power architecture, doesn't even consider mobile phones as potential applications for that architecture. Power-architecture chips these days are found in gaming consoles, telecommunications equipment, and other embedded applications, according to an IBM developer page linked from the Power.org site.
The only way IBM currently participates in the mobile phone market is by making chips for other companies that design the inner workings of the chip themselves. IBM runs a chip-manufacturing business (known as a foundry) for companies that design chips but don't have the billions of dollars required to build and maintain a modern semiconductor factory.
"We manufacture and sell customized chips to specific customers who make products that compete with the Apple iPhone. We do not 'advertise' these specific customized chips since we are dealing with a specific customer. Chips are made to the customer's specification," wrote Fred McNeese, an IBM representative, in an e-mail message.
Those chips do not appear to be the SoCs that P.A. Semi is developing; rather, they appear to be lower-level components that are needed to run modern mobile phones, such as controllers and digital television chips.
It's possible that IBM is worried about Papermaster's knowledge of IBM products or technologies that have not yet come to light. Likewise, power-management techniques that are used in the design of Power-based server processors or the components IBM manufactures could have some applications for P.A. Semi's group.
But the issue here is competition. IBM's position seems to be that even though it doesn't appear to have a single customer for the unspecified processors suitable for mobile computers, IBM is a potential SoC supplier for Apple's iPhone who could be shut out because Papermaster could improve Apple's P.A. Semi team by sharing trade secrets regarding an architecture that Apple does not appear to be planning to use.
It's a bit surprising that Apple waived its right to an evidentiary hearing that would have allowed it to challenge certain parts of the declarations filed by IBM, said John Siegal, a partner with Baker Hostetler in New York. However, that would have probably involved having to put several Apple executives on the witness stand to explain Papermaster's role and the plans of the P.A. Semi organization, and Apple is not known for its willingness to speak publicly about its future plans.
The two parties were to have discussed a schedule for "expedited discovery" and a trial at a status conference last week. It's not known what emerged from that conference; Papermaster's lawyers have declined multiple requests for comment, and IBM representatives have not commented on the conference.
AMD could finally be getting ready to explain how it intends to build chips in the future, almost a year after dropping hints that it would revamp its manufacturing strategy.
Hans Mosesmann, an analyst with Raymond James, believes AMD is about to reveal a new manufacturing strategy that will attempt to take some of the formidable costs out of making chips. Mosesmann thinks that AMD is considering spinning out its manufacturing operations as part of a joint venture with another company, making greater use of partners such as IBM and Chartered Semiconductor, or some such combination.
An AMD representative declined to comment specifically on any actions it might be considering, but said even though the company has been working on this issue for a year, no decisions have been made.
Company executives have steadily refused to explain what "asset-light" (which soon morphed into "asset-smart") actually means, but have hinted that it involves a rethinking of the way AMD develops chips. The term first surfaced during an earnings conference call in April as Chief Executive Hector Ruiz described vague, tentative plans to reduce costs by thinking of new ways to get chips built and out to customers.
There might not be a more costly undertaking in the modern technology industry than the process of designing and manufacturing high-end PC and server processors. Only six or so companies really do it in high volumes (Intel, AMD, IBM, Sun, Fujitsu, and Via). Modern chip plants cost billions of dollars to construct and maintain, and researching new leading-edge transistor designs is enormously complicated and time-consuming, and therefore expensive.
That equipment--and those suits--cost an awful lot of money, and AMD is looking to cut costs.
(Credit: Sven Doering/AMD)Even in boom times, AMD's pockets are not nearly as deep as Intel's, yet it has to spend the money to keep up. To date, it has gotten around some of the need for huge outlays by working on advanced research and development in a facility owned by IBM, using third-party chip foundries like Chartered Semiconductor and TSMC, and making the most of what capacity it does have in Dresden, Germany.
But AMD's pockets are even lighter right now. Intel has recovered from its mid-decade swoon and is chugging along without incident, while AMD has watched its fortunes spin completely around in about a year. AMD can't continue to lose money forever, and one way to cut costs is to embrace that tried-and-true 21st century management strategy: outsourcing.
Expect AMD to line up a new partner for manufacturing and/or research, expand its relationship with IBM, or accept another cash infusion from its new investor, Abu Dhabi, Mosesmann said in an interview. "They have to be competitive without having to spend billions each year to keep up with Intel," he said.
Don't expect AMD to get out of the chip manufacturing business entirely. After all, "real men have fabs." All bravado aside, there are plenty of good reasons to keep a great deal of manufacturing in-house, everything from quality control to customer flexibility.
And the thing is, AMD can only go so far. Under the terms of a licensing agreement it signed in the early 1990s (AMD licenses the x86 instruction set from Intel), AMD is prohibited from outsourcing more than 20 percent of its production.
That might not necessarily hold AMD back, Mosesmann said.
"There are some issues that have to be overcome, and AMD might be able to overcome them by saying we don't care," he said. In essence, AMD could dare Intel to sue it for violating the agreement, hoping to work out some sort of settlement further down the line when it's in better economic shape. Their lawyers talk on a regular basis already.
This is a huge, huge year for AMD's current management team. After the debacle that was 2007, Ruiz has pledged to make AMD a profitable company in the second half of 2008.
If the company has trouble making the revenue side of the ledger increase amid tough economic conditions, it's going to have to do something about the cost side, which could involve layoffs, a new manufacturing strategy, or both.
While AMD isn't giving away any information on its future fab plans, a major chip foundry is gearing up for a big new customer.
TSMC, the largest chip foundry in the world, is apparently planning to take on a new customer that wants to use the high-k dielectrics and metal gates introduced earlier this year by both Intel and IBM. Sumner Lemon of IDG News Service sat through a TSMC earnings call in which company executives mentioned a mysterious new customer would be coming on board in the second half of 2008.
There are not a lot of folks planning to use that technology at 45-nanometers in 2008. Intel is, but it's got its own manufacturing facilities. IBM is, but it also grows iits own in New York. Sun and TI are, but they've got plans to build their chips at TI's facility in Texas. Then comes AMD, which is planning to use metal gates in the future and has been thinking about relying more on foundries.
Novel technologies like metal gates--which are actually the first changes to the basic materials in a transistor since the 1960s--generally enter production on processors that are gunning for every last bit of performance. If performance isn't as much of an issue, it makes more sense to wait until the kinks are ironed out. Companies like Intel, IBM, and AMD are generally the ones pushing the envelope on both performance and new transistor technology, so it's pretty likely that TSMC's customer is someone like that.
AMD's losing a lot of money these days, and rumors have been flying that the company wants to outsource more production to foundries. TSMC already makes graphics chips for AMD. Both companies use the letter "M" in their names. Seems like a done deal.
But yesterday, AMD said it hadn't decided whether to use the metal gates on the second iteration of its 45-nanometer chips, or the first generation of its 32-nanometer chips. AMD's first 45-nanometer chips are scheduled to come out in the middle of 2008. If in fact that decision hasn't been made, then TSMC is probably looking at another customer in 2008, since AMD's 32-nanometer designs won't be ready until 2009 at the earliest.
All of this speculation comes because AMD hasn't said exactly what it wants to do to obtain future chip-making capacity. It has danced around something called "asset-light," but hasn't really explained what that will mean in the future.
- prev
- 1
- next
