Blue Diamond was a vision. I'm not talking about Viagra. The Blue Diamond decision, as I understand it, is as follows: The goal is to achieve 50% cycle time reduction and 30% cost reduction for semiconductor factories. That point on the chart is marked with a Blue Diamond. The analysis performed by ISMI showed that 300-mm Prime is not capable of achieving the Blue Diamond, so the decision is to go to 450 mm. However, subsequent analysis has shown that 450 mm is even less capable of achieving the cycle time benefits of the Blue Diamond than 300-mm Prime. Yet, the drive by ISMI for 450 mm continues. The supplier community, with the most vocal representatives coming from Applied Materials, Axcelis, ASML, SEMI, and KLA-Tencor, are demanding a reassessment of the Blue Diamond decision.

Ultimately buyer and seller agree or disagree. One supplier rightly points out that if a buyer of 450-mm equipment and a seller of 450-mm equipment agree on a purchase, then 450 mm will happen. If not, then it won't. He went on to ask, "Who was the largest equipment supplier company in 1985?" Do you know? Well, it was GCA. They are now out of business. Life went on. The industry went on. That's evolution. Are you asking me to believe that AMAT cares more about the health of the industry than about the health of AMAT? The customers do not have a responsibility to ensure that Applied Materials stays in business. Ultimately, 450 mm will happen when a buyer and seller agree on a purchase.

No one is forcing any company to invest in 450-mm. Every equipment supplier has the choice to invest or not invest in 450-mm equipment and process development. These equipment suppliers could just as easily say to the chipmakers, "We will not invest in 450-mm development until it is funded by you, the customer." The risk that the OEM faces is that a new OEM supplier will make the early risk investment and be first to market.

450 mm will reduce the number of chipmakers building state-of-the art fabs. Consolidation happens. As published recently in the industry press, the number of chipmakers that build 300-mm fabs compared to the number that will build 450-mm fabs will greatly decrease. It appears there would be four early adopters of 450 mm and another 11 or so within the next several years of 450 mm. But take a company like AMD. It has recently purchased ATI, which uses TSMC as its foundry. So AMD doesn't need to invest in a new 450-mm fab, because they can rely on TSMC to build the 450-mm fab for them. (Some say that AMD may not be able to afford to build a new 300-mm fab.) But I would assume that AMD would have a cost disadvantage of buying its chips from a foundry compared with Intel making its own chips on 450 mm.

Intel is committed to 450 mm. Intel has done its internal analysis and has determined that it needs to go to 450 mm. Intel has driven ISMI to adopt a supportive 450-mm position. It has worked to get other IC makers to invest in 450 mm and announce publicly. It has put assignees at ISMI to focus in this area. It is spending some money on development of wafers and carriers. Clearly, Intel has been consistently committed to a 450-mm transition.

Why is Intel committed to 450 mm? Perhaps Intel sees 450 mm as a way to continue cost reduction, especially as it probably cannot continue to get the same benefits from geometry shrinks that it has in the past. Perhaps Intel sees 450 mm as a method to accelerate the consolidation of the chipmakers, and Intel can outspend its competitors. Perhaps, as was seen on the 200- to 300-mm transition, the new developments in technology were largely integrated into only the 300-mm platforms. Those companies that stayed on 200 mm therefore did not realize the benefit of the newest technological innovations. This could be a key part of the strategy of that small fraternity of large chipmakers that is driving for 450 mm.