MICRO: What do you see as some of the major challenges facing the average fab manager today, if there is such a thing as an average fab manager?

POLCARI: I think the complexity of the technology. I'm sure that every generation feels that it used to be easier, right? But every generation we move down, we're beginning to get to the limits of the technology, so the complexity involved with each new node is increasing at a very high rate. So giving them the ability to manage that complexity in a fab, and some of these things you see in terms of e-manufacturing or trying to get the most out of the equipment like plug-and-play, those kinds of things, I think these are things that we need to wring every drop of productivity out of the fab. Wherever we can implement feedback control to try and reduce the complexity that's being introduced is a key feature. All of this needs to be tied to the technology as well. Obviously if we could devise technology that reduces the complexity of mask and lithography, that would be a big cost saver, and so we're beginning to look at things like maskless lithography as well. Because that brings huge cost advantages, obviously, to the product if we can implement something like that.

MICRO: There are tremendous economic pressures in terms of R&D expenditures on your member companies as well as on the equipment and materials suppliers. How can Sematech help those infrastructure companies better survive the necessary investments?

POLCARI: Traditionally, we've always worked with the supplier base in terms of funding and helping move along some projects. But I think what you see in our newer model—particularly with the ATDF, where materials and equipment suppliers can also participate now—they can take advantage of an infrastructure that's in place and not have to expend huge amounts of money on putting in, say, a process line. So it allows them to get more of an earlier look at their development efforts and how they fit in a more integrated process. Some equipment suppliers are actually putting tools on the floor in very early development stages. They can then take advantage of the rest of the infrastructure and not have to make those investments, say in cleanroom space and/or wafer processing equipment, that's parallel to what they do.

MICRO: Is it the small- to medium-sized supplier companies who can take more advantage of that? When I look at Applied Materials, they basically have their own fab. Novellus has a significant integration line. So it would seem like the next tier down would benefit the most from what you're talking about.

POLCARI: I would say they can benefit the most initially, but, I think, even long term, considering that you have IDMs that are outsourcing manufacturing, you wonder how long some of the equipment suppliers would want to stay in business with their own fab line and would think about outsourcing some of that.

MICRO: In terms of your overall view of the industry, where do you think collaboration or cooperation could be improved?

POLCARI: I think we need to do a better job working with the equipment and materials suppliers. We have some good interactions with them, but across the board that's an area that's open for opportunity. We certainly continue to work with our traditional IDMs and foundries...and continuing that collaboration is important, as well as ensuring that it's global. I think globalization really helps in terms of driving the large infrastructure pieces, say, on a lithography system. You need a consensus of the industry to drive lithography systems forward, and that's certainly one place that Sematech has always taken a leadership role, and that's something we'll continue to do and put even more emphasis on as we look forward.

MICRO: How do programs such as the AMRC and the consortium's lithography work at the University of Albany in New York work into Sematech's overall picture, and what value have they brought to the mission, to the programs?

POLCARI: The leverage we get in resources and expertise from both of these programs is an advantage to what we do. Whether that leverage is in facilities, skills, dollars in some cases, it's very important to the program to be able to broaden what goes on.

Our program in New York is particularly focused on EUV lithography, and it is, I would say, the largest EUV effort in one place in the world. It has a focal point around two pieces: one is mask blank technology, which is a critical issue, and we're doing work on reducing defects and progressing on that. The second piece will be focused around the exposure capability and learning how to use the exposure and the issues with expose and resist and the platform for resist developers. That's in place now, and we've actually done our first exposure in the last week or so. So that's progressing tremendously, and it gives us the advantage of having a very focused effort on a big problem for the industry to drive that forward.

The other piece that you mentioned is the AMRC, which is a broader type of a program, and this brings along a couple pieces to us. One is the ability to have some good fundamental science behind the programs that we're running today. So we have people looking at some of the fundamental issues that affect mobility, say, for high-k metal-gate systems. So we're doing more of the material screening and some of the fundamental work, the basic science that you need to understand. There are very few places where that's being done any more, and most of the semiconductor manufacturers do very little of it.

The second piece of it is, there's also an ability to have a pipeline of technology or future things investigated at a university level, which feeds into what goes on at Sematech. For instance, Professor Sanjay Banerjee [of the University of Texas] is doing some work related to nonclassical CMOS-type devices. His work then becomes a feedstock or a pipeline to the work that goes on at Sematech. Of course, we also look at a broader range of things, but these things are very directly tied to us and we can utilize that skill base and that look-ahead to feed things into what we do.

MICRO: There's also a level of cooperation, collaboration among consortia, such as an IMEC or LETI or an ITRI-type organization. But on the other hand, I think especially with IMEC, there would seem to be a little bit of a competitiveness to some of the work you do. So how do you work better with them, if that's necessary? And how do you deal with that kind of balancing act between needing to cooperate on the one hand and also being on a somewhat competitive plane?

POLCARI: We have regular exchange meetings with all of these entities—LETI, IMEC, SELETE. We continue to have, where it makes sense for both entities, joint programs at various levels. It all depends on what the needs are at that point in time for both organizations and what the program can look like.

In terms of the cooperation versus competition side of it, it's really just a reflection of the overall technology industry. If you look throughout the industry, you have people who cooperate on one side while in another area they're fierce competitors. Even if you look at some semiconductor companies that are cooperating on development, they actually compete on the product level. The cooperation-competition element with the consortia is really no different than what goes on in the rest of the technology industry, and it's something you learn to deal with. The thing you can't do is make general rules. You have to look at each thing individually, and there will be some cases where it makes sense to do something together for both parties and other cases where maybe it doesn't. The ones that work, and this is probably a cliché, are really the ones that benefit both organizations. If only one organization is benefiting, then a joint project is doomed to failure.

MICRO: It would seem like one of the examples of where everybody is on the same page and needs to share as much information as possible is immersion lithography. I have been struck by how far, how fast, and how much has been done. I don't think anyone has ever seen anything like this.

POLCARI: We've really progressed tremendously in immersion, and I think if you talk to people in the industry, many people will give Sematech a lot of credit for driving consensus and pushing a lot of that and funding some of the programs to look at those critical issues and make sure that there really were no roadblocks. Given all of that, though, there's still a lot more that needs to be done before we can declare success. You know, success is when you're printing wafers and manufacturing at an acceptable throughput rate and yield.

MICRO: Let's change the subject a little bit and talk about the roadmap. Obviously, it's become an essential part of the industry's progress. But again, it seems like there is some possible room for improvement. Some people have complained that there needs to be more of an economic element to the technology roadmap. How do you see that progressing?

POLCARI: I think there is an economic element just by the fact that the roadmap is really there, and part of the reason you do what you do is to gain this 25–30% productivity improvement in yield. When you look at everything tied together, even though it may not be explicitly written down on the roadmap what the costs are, the underlying theme is that you need to maintain that productivity. Maybe we need to be somewhat more explicit on that item.

We're certainly trying to put more of a cost-of-ownership flavor on some of our projects to try and flesh that out, but as you know, in the early development stages, it's very difficult to do the cost of ownership because you have to make so many assumptions. You're making a lot of assumptions and if all the assumptions work, then of course it's good, you get the right value. So what we need is to make sure we do those costs of ownership and then go verify those key points as quickly as possible....This is really what ISMI is about: looking at the cost part of the equation and making sure it doesn't get lost in what we do.

MICRO: What are some benchmarking efforts happening now?

POLCARI: One of the key activities we do in ISMI is fab benchmarking. It's an invaluable exercise for everyone, and our experience with it has always been that no matter how good you are in manufacturing, you always learn something in that benchmarking. Because not everybody is best in all categories or all items that we benchmark, we always learn something from other people.

There is a council that meets regularly to benchmark fab productivity, and they look at a series of metrics in the fab. It's very specifically looking at fab metrics, and they've come up with a set of metrics that they believe they should benchmark on a regular basis, and they look at all their fabs around the world and they do benchmarking with each other.

MICRO: It's an ongoing database or spreadsheet.

POLCARI: Yes. Then there are special ones that come up at members' request. For instance, we did one recently on power costs around the world. They all came in and shared their benchmarking, what they were paying for electricity, etc. Now that sounds very mundane, but one of our members said that there's nowhere else you can get these data. One of them actually went back to their power company and said, "You know what? Here's what everybody else is paying." And they negotiated a reduction in their power bill.

There's a whole series of metrics that, on a regular basis, the fab managers get together and actually benchmark. And our basic rule on this kind of benchmarking is that you have to share data to see data. If you're not willing to share it, then sorry, you can't look at it. In fact, what we see lately is that there's more and more interest in this [since] costs in a fab go right to the bottom line.

MICRO: What are some of the most satisfying aspects of the job for you?

POLCARI: I think that formulating a technology strategy and then actually implementing those things in the technology that have impact on what goes on. You mentioned things like immersion lithography. I think it's very satisfying that we're acknowledged as having contributed to that formulation and acceptance of immersion lithography in the industry. There are very few places you could actually say that, because we have that ability to drive together a consensus from a number of companies. So that's very satisfying.

We've also built a world-class program in advanced gate, a high-k metal gate. Over the next year or so, as we get ready for the next node, we probably have one of the best programs in the world on that and I think you'll see some really outstanding results coming out of that. The proof of the pudding is when you make those first manufacturable wafers, and when it's accepted and inserted, but I think we have an excellent program and some really excellent people. The other thing is the formulation of some of these new programs like AMRC, ATDF, and ISMI, and as they start to really take off, that's very satisfying.