by Debra Vogler, Senior Technical Editor
The next paradigm shift in the semiconductor industry business model — perhaps nudged by the growing presence of private equity — may be the emergence of a cluster alliance, or “reverse IDM” that reintegrates everything at the system level, according to Tien Wu, COO of ASE Group, speaking at the recent US-Taiwan Business Council’s event in Santa Clara, CA (Feb. 7).
Taking the audience on a ride through history, Wu described the beginning of the industry when IDMs basically did everything, including making their own process equipment, because of a lack of choice. The evolution to the fabless model came about amid the need for greater efficiencies.
The next move, to a cluster alliance, will be a kind of return to the industry’s beginning. This “reverse IDM” model, as he called it, will involve a reintegration of the segregated foundry cluster elements such that the total chain — design, manufacture, distribution channel, supplier base, etc. — is optimized at the system level.
“The foundry cluster has a problem: there’s no communication between the system houses that are the ultimate end customers,” Wu told the audience. “Today, TSMC is powerful, ASE is powerful, materials and equipment suppliers are all powerful — but none of us is willing to talk to each other.” And the people in the middle, such as the device houses, are not system houses and do not have system architecture in mind, he said.
With fickle consumers now in the IC application driver’s seat, Wu asked the audience to consider what kind of changes need to occur to adjust to this new reality. Controlling volume with just-in-time delivery, inventory changes, dealing with the numerous designs that have to pass through the pipeline, the numbers of engineers that will be required to do new designs, change of tooling, changing the manufacturing bills of materials — all will have to take place just to match a sudden change in consumer demand. “In the next evolution, whoever can figure out how to reintegrate the segregated cluster foundry elements and optimize at the system level, will win,” said Wu. “That will be the new paradigm shift.”
A portent of how the future might look for an industry based on the cluster alliance model is the iPod. Wu noted that what Apple did was figure out how to put lower-end technology together with innovative industrial design and marketing to leverage the Internet infrastructure.
A tantalizing idea deftly hinted at by Wu — while being careful to not give away his personal opinion — is the role of private equity in making the new paradigm shift come about. He observed that the semiconductor industry is now, by virtue of its size, able to command the attention of those who play on the world’s economic and political stage. [Wu’s caution was perhaps warranted, because ASE is currently mulling a $5.5 billion offer from a consortium led by The Carlyle Group.]
“When the semiconductor industry was young, and innovation was the key, it could do whatever it wanted; no one cared,” Wu observed. “But when the semiconductor industry becomes so large [that] its scale grows to the point where it affects national security and economic growth, all of a sudden every other dimension [e.g., political and social structures] will kick in.”
A prime example is private equity, he said. “Before, private equity was never interested in the semiconductor industry. Now, it is. There must be a reason for this.”
Wu further noted that, in addition to newfound availability of money and better management of investments, this paradigm shift will necessitate global management of a complicated political infrastructure, and a dedicated ecosystem. — D.V.