Reports have surfaced in recent times that fab-utilization rates for leading-edge foundry capacity are running at about 95 percent. The lack of 0.13-micron foundry capacity has extended the lead times for customers, although there are also signs of double ordering in the supply chain, said analyst Michael McConnell, who watches the foundry industry for Pacific Crest Securities Inc., based in Portland, Ore.

Still, the lead times for leading-edge, 0.13-micron foundry technology have extended from about 13-to-14 weeks in April, to about 17 weeks today, McConnell said. Lead times indicate the time from a customer orders a wafer to the time it is delivered. The lead-time for a 0.13-micron device is about "two-and-a-half days" per layer, plus "three weeks in the queue," he said. On average, a 0.13-micron device has about 35 layers.

The 0.13-micron lead times have reportedly extended at Chartered Semiconductor Manufacturing Pte. Ltd., IBM Corp.'s Microelectronics Division, Taiwan Semiconductor Manufacturing Co. Ltd., and United Microelectronics Corp.

"I think the lead times are extending across the board at select foundries," he said. "Demand is outstripping supply, which is favorable to companies like TSMC."

Indeed, after a period of sluggishness, TSMC and the other foundries are beginning to see huge demand--at least for their leading-edge processes. TSMC, for example, is seeing strong 0.13-micron demand from its core customers, such as Altera, ATI, Broadcom, Nvidia, and Motorola, McConnell said.

Despite the demand, TSMC is reluctant to expand its leading-edge fab capacity until the demand picture is less cloudy. "They know their customers are double ordering," he added.

By: DocMemory
Copyright © 2023 CST, Inc. All Rights Reserved