INDUSTRY NEWS

SEMATECH WATCH

Consortium to industry: take time to confront Y2K risks

What will happen if a small fire erupts in your fab on January 1 and you call the fire department? Will a hook and ladder rush to the scene within minutes? Or will your local firefighters be overwhelmed rescuing revelers stuck in elevators that have stopped running after midnight, leaving you alone to confront a potential catastrophe?

If that scenario sounds farfetched to you, welcome to the world of contingency planning for the Year 2000 changeover. With a little more than 200 days to go until the Big Day, experts from Sematech have been holding meetings around the U.S. to share remedies and take the temperature of the industry. SEMI has invited two presenters from the consortium to speak for two hours this month at a seminar in Beaverton, Oregon. The topic: How are semiconductor and equipment companies doing "in the race to be Y2K compliant?"

The answer is heartening. In general, says Harvey Wohlwend, chipmakers have made great strides over the past 18 months ensuring that their tools will run smoothly, not only on January 1 but also on "any number" of thorny dates, including April 9, 1999, the 99th day of the year, and September 9, 1999.

Wohlwend, Sematech's Year 2000 readiness manager, and Richard Koski, the technical director of SEMI/Sematech, will address the SEMI meeting. "We get a big audience every time we have one of these seminars," says Wohlwend. Interviewed in early April, Wohlwend and Koski had just returned from a two-day workshop in San Jose attended by 65 suppliers and 35 representatives from Sematech's member companies.

"A year ago people were concerned about getting their fab equipment ready," notes Wohlwend. "We've made tremendous progress on that." The primary concern now is how well the "public infrastructure" is prepared, he says. Hence, the need for worst-case planning.

"Harvey's program at Sematech is to make sure that Sematech members' facilities are not going to have interruptions with Y2K issues," Koski points out. Koski is focusing on risk-management concerns as an environmental, safety, and health (ESH) issue to "make sure people are putting plans in place for contingencies."

"It's my impression that a very few people have looked at the broader picture," continues Koski. "The questions I'm asking are, for instance, 'If an employee gets injured on a critical Y2K date and you ask for an ambulance, what are you going to get?' "

Would you like another example? Well, Koski notes, "a lot of traffic lights have gotten sophisticated in the last few years. . . . There's a lot of date-time information in them. Now what happens if on New Year's Eve you have 40,000 drunk drivers go home and half the signals don't work?"

Koski says each device maker needs to check with the local municipalities and ensure that, if the powers-that-be aren't prepared for potential Y2K contingencies, the semiconductor manufacturer has covered its bases. "Call up city officials and ask them, 'If I call you on this date what am I going to get? If I have a fire, is that my problem? Are you going to come help me?' Or what if you have a problem with your acid waste discharging pH2 acid [into the local waste system]? My talk is designed to encourage people to start a dialogue among businesses in the local area and among business and community leaders."

Koski notes the "governmental people and business risk management people" to whom he has spoken "are in varying levels of preparedness, and I think the questions I ask in this talk tend to increase the level of dialogue."

However, the Sematech expert has spotted a worrying trend related to potential legal liability. "My perception is that we've gotten to that critical juncture where people are a long way along on their programs, and they're starting to get worried about lawyers. My largest concern right now is that we're really close to critical mass and people are starting to feel comfortable with their programs, and this issue of reliability and culpability is starting to raise its ugly head. Everybody is starting to shut up when they should be starting to talk."

Koski maintains that the equipment segment of the semiconductor industry has prepared well for the Year 2000 problem, however. "Designing failure scenarios into equipment has paid us back in spades."

Koski was asked whether it would take just one major tragedy to wake up the industry to the potential pitfalls lurking after midnight, December 31. "It wouldn't have to be that big," he responds. "If you can't get emergency responders there, a small fire could turn into a big one and then it's a billion dollar problem."

The industry has shown increasing interest in one area in particular, according to Koski. "There is a large concern about the transportation infrastructure," he notes. Chipmakers are worried about "getting materials from point A to point B over the highways because the gas stations may not work or signals may fail." As a result, "a lot of fabs are stockpiling materials out there. And some of it is not the kind of materials we like to stockpile. Are we adequately managing that? I think the answer to that is 'yes,' but there's a lot of surprise when I ask that question."

Rick Scott, who brought Koski into the Y2K program at SEMI/Sematech two years ago, warns that inventory hoarding has a potential downside. "With any contingency plan you put in place, if that contingency doesn't occur you're going to have to undo it after all this passes. An example of that would be bringing in a bunch of inventory early. If you do that, you'll have to work that inventory off. And that creates a slump in purchases of that inventory for a time."

Referring in particular to raw materials, Scott says the plans "can echo around the industry for quite a long time." He notes wryly that such millennial planning "is sort of the opposite of everything we've learned about resources and inventory." The anti-JIT movement, as it were.

"That's one of the things that the industry is now focusing on," Scott elaborates. "The industry doesn't want to set in place contingency plans that will have to be undone at great expense and echo around the industry for a long time."

Having said all that, Scott praises the consortium's efforts to help the industry prepare for the Year 2000 problem. Chipmakers have devoted a lot of time to preparation, "which tells me that manufacturers have moved to the next place. They're now fairly confident that either the tools will be ready or they will have workarounds in place. They have now moved on to contingency planning, which means getting ready for the unforeseen."

With contingency planning on the front burner, manufacturers and suppliers are turning their attention to "a phenomenon called the 'systemic cumulative effect,' which is another very sophisticated concern," says Scott. "A real simple example of that would be, what if we've missed something about how some of our really important tools such as steppers are affected? And rather than have one stepper go down or one implanter go down—which is not uncommon in our industry—a lot of them go down and something systemic occurs. Then we have a problem larger than the particular type of problem we handle day in and day out."

This type of concentrated gaze at the Year 2000 problem and potential cumulative failures "tells even the uninformed observer that the industry has become very sophisticated in how it's looking at this problem," asserts Scott, Sematech's director of software programs. "That's a very good thing. They're way past panic and deep into engineering the problem, taking it apart and looking at all the possibilities."

Is all this costing chipmakers and suppliers, um, a lot? Scott laughs good-naturedly. "It's costing a lot. You can quote me on that: Rick Scott says, 'It's costing a lot.' In fact, you can say 'a whole lot.' " He quickly notes that no company, of course, wants to share publicly just how much it's spending on all this head-scratching, stockpiling, and planning.

In any case, many companies are logging on to Sematech's Web site to look at the consortium's Y2K readiness testing program. Sematech, says Wohlwend, "has defined a series of test scenarios as a minimum set" for Year 2000 preparedness. The site has received "lots of hits from all over the world," Wohlwend notes.

Remediation efforts have been successful then? The Sematech readiness expert notes that on the potentially problem-causing date of January 1, 1999, some companies experienced difficulties. "Fortunately, they didn't happen at our fabs," Wohlwend notes proudly. "I did a survey of our member companies and not one fab had a problem."

You wouldn't need to scratch your head very long to figure out the disastrous effect a Y2K-caused problem would have on yields. Notes Koski: "The Y2K problems potentially impact the infrastructure—areas like water delivery, chemical delivery, air control, abatement—those types of things. If any one of those screw up, they can potentially impact yield."

But it is the area of ESH that concerns Koski. "Sematech is focused very much on equipment. But no industrywide organization has focused on the infrastructure; that is, what goes on outside the cleanroom. There's potentially a large impact there."

Sematech's Year 2000 Problem page can be reached by logging on to http://www.sematech.org/semi-sematech.

Illustration by James Schlesinger

MicroHome | Search | Current Issue | MicroArchives
Buyers Guide | Media Kit