There's plenty of news about AMD's plan to split itself into two. One part will design chips while the other becomes a joint venture with investors from Abu Dhabi. But this isn't a story just about AMD. Indirectly it's about the entire semiconductor industry and the even bigger implications for the global economy when the companies that power everything from greeting cards to autos and appliances can't figure out how to make money. If they can't stay in business, how is anyone going to make anything?
Most attention is being focused at the moment on the dollars flowing into beleaguered AMD and the creation of a new venture, temporarily called Foundry Co.:
In conjunction with the spin off, Abu Dhabi's investment arm, Mubadala Development Co., will invest $314 million to more than double its current stake in AMD to 19.3% from 8.1%.Another entity backed by the Persian Gulf state, Advanced Technology Investment Co., will invest $2.1 billion for a stake in Foundry Co., which also will assume about $1.2 billion of AMD's existing debt.There's a lot of positive perception of the move. Given where AMD has been, that's not surprising. Almost anything would have been a financial improvement.
Advanced Technology Investment then plans to contribute between $3.6 billion and $6 billion to Foundry Co. over the next five years to fund the expansion of the company's chip-making capacity. That plan includes the construction of a new facility in Saratoga County, New York.
But for a moment, look beyond AMD's acquisition of graphics chip maker ATI. Look beyond competition with Intel on one front and Nvidia on the other. The entire industry is in pain and trying to figure out how to keep making money, not because of acquisitions, but because costs have become insane.
According to analyst firm iSuppli, semiconductor quarterly net profits are now in the single digit range from 17 to 19 percent even as recently as 2004.
"To a degree, conditions in the semiconductor industry have been impacted by short-term events, such as the market volatility in 2006 due to inventory write-offs and price wars in major product segments like DRAMs and microprocessors," [iSuppli CEO Derek Lidow] observed. "However, the long-term trend indicates that the semiconductor industry--which historically has been good at capturing profits in the electronics value chain--seems to have lost its money-making touch."I recently spoke with Cypress Semiconductor CEO T.J. Rodgers. He said that his company's development costs for a new chip can run anywhere from $5 million to $100 million for the most complex models.
That's not that unusual. It can easily run tens of millions of dollars for each development iteration of a new chip through prototype manufacturing, and that's not even using the cutting edge geometries. New fab plants cost several billion dollars and need regular infusions of heavy capital investment. Some critics seem to see the manufacturing end as almost a distraction. Real semiconductor companies are supposed to focus on design.
The problem is that the split between design and manufacturing, which seemed so natural in the 1990s, has left the chip companies struggling to figure out how to design for the new technologies. The industry has recognized the fundamental issue. Either semiconductor companies find ways to design chips that don't need multiple expensive prototyping iterations or many will have to give up the cutting edge for the predictable.
Unfortunately, customers won't wait. Even if consumers would, companies look to semiconductors to provide the additional features they want to include to create product differentiation and shore up margins on premium products.
The problem facing the global economy is not that semiconductors by themselves are an overwhelming percentage of GDP. However, too many industries -- automotive, healthcare, all manufacturing, for example -- depend on the existence of semiconductors. If this part of high tech tanks, it won't be long before virtually everything does. Think the credit crisis is bad? Imagine companies being unable to make, oh, just about anything with a high tech aspect for a reasonable amount of money. Better hope that someone figures out how to make the next generation of chips at a reasonable price.