SAN FRANCISCO—An executive panel discussion at the Semicon West fab tool vendor trade show here Wednesday (July 13) revealed widespread agreement on the strength of the semiconductor equipment and materials industry's business prospects over the next several years as well as some uncertainty on the pending technology transitions to extreme ultraviolet (EUV) lithography and 450-mm wafers.
EUV has for years been considered the heir apparent to optical lithography. But EUV has been pushed out several times, and a number of issues remain unsolved, including lack of an adequate light source and the absence of a pellicle for EUV photomasks. Some economic issues remain sticking points, including the expected cost of production tools—reportedly as much as $125 million each—and throughput.
The move to 450-mm wafers, which was initially resisted by tool vendors, now appears to be a foregone conclusion, although the previous target date for production on 450-mm wafers in 2012 will almost certainly be pushed out.
"The challenge for all of these technologies is that we cannot decouple them from the economics associated with the duty they are performing," said Rick Wallace, president and CEO of KLA-Tencor Corp. "An EUV tool that runs 20 wafers per hour has far less use than a tool that runs 150."
Steve Newberry, CEO of Lam Research Corp., echoed Wallace's comments, saying that whether the move to 450-mm happens at all will be determined over the next three years as chip and equipment vendors study the economic feasibility of the move. He also voiced concern that in three to five years—the soonest he predicts that 450-mm could be put into production—the semiconductor industry will be grappling with putting a host of other new technologies into production, including 3-D NAND flash devices, vertical DRAMs, FinFETs and possibly MRAM memory and the use of 3-5 materials.
"If you think about all of that complexity, and going to 10-nm, do you really want to throw 450 on top of that?" Newberry asked.
"The economics have to make sense," Wallace said, adding that even if they don't make sense some equipment vendors may push 450-mm for the purposes of gaining market share.
"We've seen that movie before, too." Wallace added that funding for the move to 450-mm must come from other sources besides the equipment vendors.
"The [equipment] vendors had to pay for an awful lot of implementation of the next generation of scaling," said Terry Brewer, founder and president of process and materials supplier Brewer Science Inc. "At 300, it was a touch and go question. At 450, I think it has to be revisited. I think it challenges the entire idea of Moore's Law and [whether] this linear scaling process is a good idea. It's certainly a good idea for somebody. But it's not a good idea for someone that has to pay the bill."
Once 450-mm tools become available—some have been announced this week—equipment suppliers won't have a choice about whether to follow the crowd and supply 450-mm tools. "If you don't go [to 450-mm] and it gets implemented, you are gone," Newberry said. "The challenge for equipment suppliers is to collaborate so we don't have five early pilot lines that drive up the R&D costs. We are going to go. We will have 450-mm. What I don't know is when it will be implemented."
Wallace added that when 450-mm tools are developed, he expects a number of chip companies to migrate to 450. Some analysts have speculated that only the deep pocketed top three to six chip makers in the world will ever move to 450. "The hard one is the first one," Wallace said. "Once that happens, more companies will adopt 450 than just the three to six envisoned. It's the getting from here to there that is the sticking point."
Next: EUV, business conditionsSemicon West, Steve Newberry, Rick Wallace, 450, EUV Semi West exec panel questions 450-mm, EUV
No comments:
Post a Comment