Chip Venture Lacked Money and Customers

For Sanford Kane, the end of U.S. Memories came swiftly and without any advance warning last Wednesday in a Dallas hotel room.

After meeting all morning with representatives of 11 electronics firms--seven that had already agreed to support the novel chip-making cooperative and four others that had been straddling the fence--Kane told the group that it was time to finally count how much money everyone was willing to give the venture.

With that, “the room got real quiet, real fast,” Kane, U.S. Memories’ president, recalled Monday after announcing its demise.

Wilfred Corrigan, the semiconductor industry leader who had agreed to serve as chairman of U.S. Memories, tacked up a big poster board listing the potential backers and two empty columns next to their names. One column was for the amount of money that each company was willing to invest in the venture; the other for the amount of product that the companies were willing to purchase from the manufacturing cooperative.

As the numbers were posted and tallied over the next hour, Kane said it became all too obvious that U.S. Memories had neither the minimum financial backing nor the basic level of guaranteed customer support it required to open its doors.

“Believe me, I had absolutely no indication that it would turn out this way,” Kane said Monday. “I walked into that meeting with as much optimism as ever.”

In the end, Kane said, the assembled group, which included some of the largest and best-financed members of the electronics industry, could not collectively commit to raising the minimum $150 million that he said the venture needed immediately to begin operating.

Further, he said, with the exception of International Business Machines and Digital Equipment, the computer makers present would not commit to buying a significant portion of their requirements for dynamic random access memory (DRAM) chips from the cooperative.

Without either, the venture was dead, Kane said.

In hindsight, some might question why Kane was so surprised at the outcome. In seven months of pitching the plan for the manufacturing cooperative, a good deal of the industry response that he received was negative.

Although the proposal was rejected outright by a handful of companies--notably Apple Computer and Sun Microsystems--several other companies questioned whether the plan was overly optimistic and had accurately gauged the ability of a start-up company to supply reliable state-of-the-art technology, both now and in the future.

In all, Kane said, about two dozen companies reviewed the $1-billion business plan that he had prepared last fall, a step he didn’t even think would be necessary when the proposal was unveiled last June.

But even a formal business plan failed to satisfy some. Although Sun and Apple made their decisions to stay out of the venture without ever seeing the plan, Kane said other companies evaluated the plan until the very end, without making a firm commitment one way or the other.

“The whole process has been extremely frustrating,” Kane said.

Initially, the business plan called for the venture’s partners to invest $500 million, with an additional $500 million to be raised from debt securities to be sold on Wall Street. The proceeds were to be used to fund the opening of three semiconductor wafer manufacturing plants.

However, after potential investors complained that the plan was overly optimistic and grandiose, Kane scaled it back, offering to raise just $150 million from partners, with an additional $200 million to come from Wall Street. And the company initially was to open just one manufacturing plant.

As a trade-off for the lower level of industry investment and infusion of equity from Wall Street, Kane said, at least 75% of the venture’s products had to have guaranteed buyers before the company could begin operating.

But even a greatly clipped U.S. Memories apparently couldn’t fly.

“I wish it could have been otherwise,” said William Warwisk, president of AT&T;’s microelectronics unit, which was considering whether to join the venture. “Our competitiveness would have been better served if U.S. Memories had become a reality. But the culprits were a rising supply of DRAMs and falling prices for them.”