TWST: Could you tell me a little bit about the sectors that you cover, and the companies that you cover within those sectors?

Mr. Niles: I cover two main areas, one being computer hardware, so it's the traditional IBM (IBM), Compaq (CPQ), Dell (DELL), etc. in that area. And then I cover another area, which is what we're having this conference for, which is semiconductors. Within semiconductors, I deal primarily with the larger capitalization, broader end-market IC suppliers. This includes companies such as Intel (INTC), Texas Instruments (TXN), LSI Logic (LSI), Altera (ALTR), Xilinx (XLNX), Microchip (MCHP), Micron Technology (MU), etc. So those are the two broadly defined areas that I currently cover. There's a third research focus that I'm starting, which is called digital media infrastructure, which includes the systems companies as well as the semiconductor IC companies, which will sell into this area. And this is an area that includes consumer electronics devices, Internet/information appliances, set-top boxes, etc. and the semiconductor companies that sell into these end-markets. So that's another area that we're starting to focus our research on, which is more of a new and emerging area. One of the recent reports that we wrote, called the 'Internet Appliances Report,' touches on this.

TWST: What's the investor sentiment in the second quarter of 1999 towards the semiconductor sector?

Mr. Niles: The way we view it is that investors are cautiously optimistic. We have two really different sectors, I think, within the semiconductor space. We have the companies that are focused on PC- related products, and then we have the companies that are focused on the non-PC-related products. And the non-PC-related companies are all doing very well. The guys who are focused just on PC-related products for the most part, like an Intel or a Micron, are obviously having a much more difficult time, and probably will continue to have a difficult time as we go through the end of this year. So it's really a split between those two areas in how investors should be viewing it.

TWST: How does the non-PC sector divide up? What kind of products, what kinds of end-user markets, and what are the factors driving the non-PC- related products?

Mr. Niles: You have some broad areas. You have the consumer electronics area, such as digital cameras, set-top boxes, DVD. You've also got automotive. You've got industrial. And you have communication, which includes telecom and data communications. All of those areas, in one way or another, especially in the consumer electronics and communications areas, are seeing fairly strong demand. Demand for networking equipment and telecommun-ications equipment is very strong. There are more and more electronics going into cars and industrial applications. Consumer electronics devices, like the Sony PlayStation and the Nintendo64, are also absorbing a lot more chips. So those are some of the other areas that are seeing pretty good demand coming that haven't traditionally been as important as PCs.

TWST: Why is this sector doing so much better than the PC? Is the demand environment better, less price pressure on the OEMs? Or is it less obviously competitive from a supply point of view?

Mr. Niles: I think a couple of things, and you kind of touched on them in your question. The PC market right now is going through its own cycle where prices are coming down very rapidly for the PCs themselves. And as result the business model in the PC market is changing. You're going to more of what I call a cell phone model. You get the PC for a $199 price point, but what people are really trying to sell you is that $19.95-a- month Internet access fee. That recurring revenue stream is what they're really after, and they're using the PC as a mechanism to get their hands on that $19.95 a month. So if you're a pure hardware supplier, that makes life obviously difficult because you're not getting a part of that $19.95 a month. You're just getting the revenues from the hardware sales. That makes the PC market very difficult and challenging in the near term. If you look at the telecom, networking, or other markets, those markets aren't going through a big pricing trend downward, so it's a little bit easier to make money selling into those markets, and that's what you're seeing right now. There's the big macro trend going on in PCs. And PC is still a high-volume market and still the biggest market, but in the near term you just have some issues associated with it.

TWST: In the PC market, is there a sense that processing power is no longer the first order of constraint on overall system performance? In other words, nobody is actually waiting for the next version of the Pentium. Is there any of that going on?

Mr. Niles: I think there's definitely some of that, but it also has to do with the other two legs, which is software and bandwidth. On the software side, you haven't had a lot of new stuff that has necessarily taxed the processor. Microsoft (MSFT) has finally gone ahead and done some of its new software releases, which are going to take a little bit more processing power, so that will be helpful. But the final piece is bandwidth. You have these very powerful PCs, but what a lot of people are buying PCs for now is getting access to the Internet. Well, if you've got a 56-kbps modem and you're trying to download video clips, the modem is your real problem. It's not your PC or your processor. Now once the bandwidth constraints get fixed, and you get DSL or cable modem deployment, and the information coming to your PC switches from 56 kilobits per second to 10 megabits per second, then you're going to tax the processor again. And the good news is that a lot of that stuff is coming in 2000. So even though the rest of this year I think will be fairly difficult for companies that primarily service the PC industry, I think next year will actually be pretty good because as the bandwidth constraints start to lift, then once again it will go back to your processor being the bottleneck, and not your bandwidth.

TWST: Are there any major technology advances within the semiconductor industry that have allowed any companies or sectors to push ahead of the pack, or develop a competitive advantage that makes them more profitable?

Mr. Niles: I'm not sure that there's anything radically new. What I would say is that there have been some more incremental developments that have happened. I think you've had both MIPS (MIPS) and ARM (ARMHY) come out and show the power of the intellectual property model as it applies to the microprocessor segment. And that's something that Rambus (RMBS) showed earlier, so that's kind of extending what has already happened in that area. I think as you look at some of the trends in the logic market there's much more of a focus on integrating the digital and analog pieces together, with different companies trying to have both sides of the equation. You know, LSI Logic has done that, and Microchip has used its mixed signal expertise and built a line of mixed signal products that work with their microcontrollers. Texas Instruments is doing the same thing by building mixed signal products that others have typically provided that work with TI DSPs. So you're seeing a lot of the bigger logic companies starting to pull the mixed signal pieces that make sense, and incorporate them into their own product strategies. So that's another small trend you're seeing. You're also seeing increasing outsourcing to foundries. The amount of money that it takes to build a brand new fab is so enormous that a lot of companies are turning to outside foundries, like TSMC (Taiwan Semiconductor, TSM) or UMC, to provide additional capacity for them going forward. Those companies that are outsourcing are trying to focus more on the design and sales as opposed to the manufacturing. This trend towards outsourcing will only accelerate as you move to 300-millimeter fabs and .15 microns. The amount of volume that you need to run through the fab to make it economical will more than quadruple from current levels. As fabs go from $1 billion a pop to more like $2 billion a pop, the trend to go ahead and partner yourself with an outside foundry will only increase. We saw this with LSI partnering with Silterra recently. So those are trends that have been to some extent happening, but are becoming increasingly prevalent.

TWST: For the companies that focus more on design, do they ever get ahead of the fabricating companies, design something that nobody can build because they've gotten ahead of the curve?

Mr. Niles: No, typically it's the other way around. You can build a silicon die a lot bigger than you can actually design one because to some extent, the design tools that the EDA vendors have been providing have been notoriously bad in terms of keeping up. Your manufacturing has advanced a lot further than your design capabilities have. So typically it's not a problem that you can't build it. Most of the time it's that you can't take advantage of the process technology you have, which is big opportunities to vendors like Synopsys (SNPS) and Cadence (CDN) and others in the EDA space.

TWST: What is the overall supply capacity environment for the semiconductor business? I know that for the last several years there has been an overcapacity in the industry. How is that evolving?

Mr. Niles: I think, again, it depends on who you're talking to. The DRAM vendors are still in an oversupply situation and DRam prices continue to fall at a faster than historical rate. In the microprocessor world National (NSM) obviously got out of its Cyrix microprocessor business. Advanced Micro (AMD) this last quarter produced 6 million units and sold less than 4 million, and Intel has got more than enough capacity right now. The processor vendors in total have plenty of capacity at this point. However, if you switch over to programmable logic vendors, ASIC vendors, or the non-PC-related guys, a lot of them are struggling to keep up with demand. So it really depends on who you're talking to more than anything else. It's really interesting because I think this is the first time you've seen a very noticeable split in terms of the semiconductor industry. Usually, for a combination of reasons, they seem to go together, but this time there seems to be a definite break in that pattern.

TWST: Given this background, which are the companies that you're telling investors are well-positioned right now, in terms of the sectors they're targeting, in terms of the supply/demand capacities that their sectors are facing?

Mr. Niles: LSI Logic, Microchip, Texas Instruments, SanDisk (SNDK), MIPS, Altera, Xilinx and Cypress (CY) are names that come off the top of my head.

TWST: Are there any standouts, any case studies within that group that you would like to highlight?

Mr. Niles: Yes, I think LSI, Texas Instruments, MIPS and even Xilinx are probably interesting in different ways. MIPS is very interesting because of the intellectual property model that it has. It is helping to pave the way into a whole new area where you make money as a semiconductor company, not by building the chips, and not even necessarily by selling the chips, but by designing them, licensing, and collecting royalties. It is also a great play on the whole Windows CE market and embedded processor market in other words, embedding intelligence in a lot of different electronic devices. Texas Instruments is a great play off of the whole wave towards digital signal processors. The more you need to manipulate data quickly, transfer it quickly, compress it, decompress it, or just move data around in a very fast manner, the more you're turning towards digital signal processors, and TI is the market leader in that. It has almost half the market, and that market share will probably continue to grow. So it's sort of what I call the second processor company behind Intel, but its market is growing much more rapidly than Intel's is right now because it serves a very diverse set of markets outside the PC market. The third one, LSI Logic, you not only benefit from its exposure to telecommunications and data communications, you also have a really good play off the whole consumer electronics market. They have very good relationships with a lot of the big vendors that sell digital cameras, DVD players and set-top boxes. So there you're getting a play on some markets that are very big and very fast-growing as you go towards the second half of the year, when those consumer markets really should start to accelerate. There is also Xilinx. It is the leader in the whole programmable logic space with its new Virtex products. They are really helping companies get to market faster, so they're enabling time to market. Those are four standouts that you could pick. Microchip is another one on the theme of embedding intelligence in a lot of different devices, much like we talked about with MIPS. Microchip provides these chips that cost a couple of dollars, that you can put into washing machines, toasters, watches and pagers. They basically enable you to embed intelligence in any device no matter how low the cost.

TWST: You've written about Internet appliances and consumer electronics as they relate to the data processing world and PCs. Can you talk a little bit about that market? What are the main segments that fall into that category, and which of them do you think are ready for prime time?

Mr. Niles: The Internet or information appliance market, however you want to call it, is a new market in the sense that it's really the convergence of a bunch of different areas. It's the convergence of PC hardware, consumer electronics, the Internet, advertising and media. So these Internet appliance devices like a Web TV, what is that? Is that a PC? Well, not really. Is it a consumer device? Yes, it is, but it has got portions of the Internet in it, and it's also got advertising. It's a new product. The new set-top boxes are kind of in that category. The new game machines that are coming out from Sony (SNE) and Nintendo (NTDOY) will be able to access the Internet. They sort of fall into that category as well. Even the U.S. Robotics PalmPilot now has wireless access to the Internet. There are going to be a lot more devices like these that take computing capabilities, wireless connectivity, access to the Internet, and have some piece of advertising or e-commerce associated that create this brand-new set of what I call information/Internet appliances. Smart phones are another example. And the semiconductor devices that you need to serve them will have some of the same characteristics. You're always going to need a processor, and you're going to need some memory. But you're going to need a much more complex ASIC, much more focus on software compatibility with hardware and more focus on power consumption. There are going to be a lot of different things relating to the silicon, and a lot of trade-offs that you maybe haven't had to deal with before. Obviously, the biggest problem people have with their portables today is that you can't take them on a flight from the East Coast to the West Coast without your batteries dying. On the flip side, how much performance are you willing to sacrifice to get the longer battery life? What about the weight added by the components for wireless connectivity? So there are a lot of pieces to it, and some of the companies that come out to deal with that may be different from the ones we're seeing today. The other part of it is that the electronics companies that make a lot of money off these IA products may be also somewhat different. U.S. Robotics did a great job with the PalmPilot, almost single-handedly driving the whole PDA market. All of these devices are going to need good visual interface. Flat panel displays are going to take off. There is some talk about things called personal recorders coming out, which are sort of a hybrid between a VCR and a computer. They help you manipulate television signals that are coming in to create your own virtual television network. So there are a lot of different devices coming, and I think you'll see them play out over the next year or two.

TWST: What I hear you describe is almost a monumental fragmentation of the appliance market, which perhaps crudely could have been described as a Nintendo box and a PC five years ago, and now you're describing a vast array of potential products. As an analyst, how are you seeing consumers starting to make those choices, and how are you divining which companies that are attacking that market are going to succeed?

Mr. Niles: With a lot of legwork is the best way to put it. It's a difficult task. Because things are changing so rapidly you're constantly trying to figure out what is the next step. For example, America Online (AOL) just announced a deal where it's going to be subsidizing the sale of a PC. Well, are they really helping to sell a PC, or are they viewing it as more of an Internet connection device? What is that, and how is that going to evolve in the future? Are you going to add video capability so that it's more of a videophone? AT&T (T) has talked about what it's going to try to do in that space. So there's really no easy answer to the question. One of the things I try to do, with Keith Benjamin and Mike Graham, who are our two Internet analysts, is spend a lot of time talking. We also visit some companies together because it is really a mixture between several different research focuses. Arnab Chanda, who used to focus on semiconductors and computing systems with me, is now primarily focused on this whole new digital media infrastructure space. We all rely on Arnab to help ferret out how the marketplace is changing because the other thing is, as you know, a lot of companies have some really good ideas, and then it comes down to who can actually execute. The Apple Newton is a great example of a lot of hype and very little follow-through. So there's no easy answer to determining the winners and losers. It's basically going out there, visiting a lot of companies, and to some extent trying to figure out what you as a consumer would be interested in buying, because I consider myself to be a pretty typical consumer. I own a Sony PlayStation as well as a Nintendo64. I have a cell phone, a PC, DVD player and HDTV. I like electronics gadgets, but I don't buy all of them. I didn't buy a Newton, for example. So I think there's no simple answer. It's just a lot of work. Also, I like to talk to analysts in other areas. AOL's approach to the information appliance is going to be lot different from Compaq's approach to it. So it's trying to look at it from a bunch of different angles and figure out who has the best way of skinning the cat.

TWST: If we're talking about consumer appliances, often presumably less expensive consumer appliances, is this market going to play into the hands of the classical consumer electronic companies, like the Sonys and the Matsushitas?

Mr. Niles: My feeling is that to some extent it will, because a lot of these devices are not PCs. And the PC companies have a very PC-centric view of the world. The thought of having to go from selling $2,300 PCs to selling $200 PalmPilot knockoffs is a pretty disturbing model if you're Dell. However, if you're a Sony, all of a sudden that's not a disturbing model at all. It's just another consumer electronics device that you're selling. So to my way of thinking, the consumer electronics companies have an advantage, because they're very focused on what the customer wants to buy, and not necessarily as much on the semiconductor technology or the electronics behind it. They're trying to figure out what the needs of the customer are, and they're trying to come up with a product that fits them. In the PC industry there haven't been a lot of monumental changes. The box pretty much looks the same. Capabilities haven't changed much. It's just sort of the speed at which you do things. The PC has made some incremental changes, but there has not been anything recently that I would say was revolutionary. You still can do spread sheets. You still do word processing. For the most part people do not do a lot of telephone calls through the PC. The biggest new thing for the PC has, quite honestly, been accessing the Internet. But the PC is still not necessarily the easiest thing to use. Well, the consumer electronics companies are coming at it from a different angle. They're saying there are 900 million people, roughly, who have TV sets. There are about 300 million people, roughly, who have PCs. There are about 100 million who use the Internet. They're coming in from the standpoint of, how do we make whatever this appliance is easy enough and cheap enough so that the 900 million people who have a TV will want one of them. That's the approach they're taking. 'How do I get it below $200, down to a TV price point? How do I make it easy enough to use so that everybody will want one?' So it's a slightly different mind-set from potentially what the PC vendors are coming at it from.

TWST: Getting back to the semiconductor manufacturers serving those markets, is that going to put increasing pressure on the manufacturers that are primarily targeting the PC markets?

Mr. Niles: Oh, sure. Don't get me wrong, the PC market is still the biggest market for semiconductors, so you still have to focus on it. But in terms of where things are growing fastest, at least right now, it's outside that area. I think 2000 is going to see a resurging boom in the sales of PCs as bandwidth issues start to lift and people start to demand faster processors. So this is not the death of the PC industry. It's just sort of a pause for a period of time, with other areas becoming increasingly important. So I think from that standpoint you still need to be very focused on the PC industry. You're still selling 100 million PCs a year and growing. So that's very important. I think what's different, though, is that you may need to think about what other computing devices that aren't PCs are demanding. Is it more important to have performance, or is it more important to get those connectivity features in? Or what about power consumption? Maybe I need to really focus more on software interoperability? Does this run Windows CE? Does it need to run Windows CE? What's the display characteristic? Can I get away with the black-and-white display? Maybe I need a color flat panel screen. Does it need to be an interactive screen? There are a lot more trade-offs and system features that the semiconductor companies are going to have to focus on. So it's not going to be just focusing on the design of the chip. You're going to have to focus on how that chip works with the software out there today, and with the connectivity standards that are evolving. There are going to be a lot more trade-offs. So the designer is going to have to spend a lot more time talking to the software engineer, who is in turn going to be talking to the wireless expert. There's going to be a need for a lot more systems knowledge than there was before.

TWST: Is that in any way going to impact who will succeed, and who will not succeed?

Mr. Niles: Yes. The bigger semiconductor companies, like LSI, Intel, and Texas Instruments should have a leg up. They have enormous capital resources, access to intellectual property, can purchase or license what they don't have, and have a broader view because they're just bigger and play in a lot of different markets. If you're trying to take a more systems-oriented approach to this market, you're probably better off if you don't have just a networking focus, wireless focus or mixed signal focus. I think that if you have a little broader base of knowledge, it may enable you to compete more successfully because you can adapt a little bit more quickly and pull together different pieces of technology that may be useful in these newer products.

TWST: Can you give me just a few words about the semiconductor conference and its goals?

Mr. Niles: I think the semiconductor conference happens at an interesting time because it's at the end of July. Typically the summer has been a major inflection point for this industry, with things either getting a lot better or a lot worse in every year that I can remember being involved in it. I think this could very easily happen again. Another goal of the conference is to try to sort through the emerging trends, including some of the ones we've talked about such as the opportunity in information appliances, and others, like whether there is a big change in the PC market and how you're going to make money supplying to it. There will also be a big focus on the back half of the year, when a lot of our companies make a majority of their money. You have the big PC cycle in the back half of the year related to the holidays and back-to-school. Is the rapid decline in PC ASPs going to change the profit picture for the semiconductor companies? So I think those are the major themes that we're going to try to focus on.

TWST: One last question are all the Y2K issues embedded in the semiconductor PC markets understood and discounted for now, or are there any issues that you think have not been fully understood that are going to reveal themselves?

Mr. Niles: My personal feeling is that Y2K is still not completely discounted or understood. My feeling is that you're going to definitely have spending lockdowns as you go through the back half of the year. There's no way around it. A lot of companies have tried very hard to get themselves Y2K-compliant, so they spent much more money than they would have otherwise getting to that point over the last year. Once you've got your systems debugged, the last thing you want to do is introduce any variables into it that could potentially mess it up. So I think there are going to be some effects, and we're going to see them in the back half of the year, from companies that have spent a lot of money, and pulled in some of the spending in the first nine months, shutting it down in Q4. And I don't think people have accounted for that because it's very hard to quantify how much that's going to be. I know internally, for example, we're not going to be upgrading a lot of our PC software until we get to the first quarter of next year. And that's just one example. A lot of companies I've talked to are in a similar situation. They're not going to be upgrading their PCs or their software till 2000. There are a lot of different variations on this, where they've been trying very hard to get to a point where they feel comfortable. So I think the turn of the millennium is one of the big wild cards that we've haven't had to deal with for the last 1,000 years and I don't think the Egyptians had PCs. That, unfortunately, creates some real tension because the one thing the stock market hates is uncertainty. If you know it's going to be bad, that's viewed as better than not knowing. And I think that's one of the things we're going to have to deal with that's going to make the back half of this year really challenging not knowing what is fundamental demand and what is Y2K- related. I've said this in several speeches I've given before, but I think the back half of the year is going to be one of whipsaws. You're going to have very sharp rallies, and very sharp corrections and very sharp rallies. And I think it's going to be a seesaw battle back and forth because I think you do have all these different trends playing out at the same time. One is that typically it has been a good time to buy these stocks exiting the summer. However, let's say you buy the stocks in August, right in front of a Y2K slowdown. That's probably going to be a big problem. So you have a lot of interesting big macro trends happening all at the same time that could make this one of the trickiest second halves that we've had to deal with ever.

TWST: That should also make it an interesting conference.

Mr. Niles: Oh yes, I'm really looking forward to it. It's unfortunate almost, that we have to be so involved in putting the conference on. It would be just fun to sit back and listen to as many companies as I could.

TWST: Thank you.

DANIEL T. NILES BancBoston Robertson Stephens Bank of America Building 555 California Street San Francisco, CA 94106 (415) 693-3241