한국   대만   중국   일본 
Shawn and Kim Hailey
The Wayback Machine - https://web.archive.org/web/20131111155525/http://silicongenesis.stanford.edu/transcripts/hailey.htm
Interview with Shawn and Kim Hailey
December 29, 1997
Hanalei, Hawaii

RW: Today I'm with Shawn and Kim Hailey, the founders of Meta Software. Shawn and Kim developed and marketed HSPICE, the industry standard IC circuit simulator of the 1980's and '90's. We caught up with them on a stormy day at Shawn's winter retreat in Hanalei, Kauai, Hawaii. That's Shawn on the left, Kim on the right. So guys, where'd you go to school and where were you born and things like that?

KH: We started out in Lubbock, Texas, went to Texas Tech, and even though math and chemistry seemed to be what we took the most courses in, we had a number of side projects in electronics. And Shawn started a little company, Datalink. Do you want to...

SH: Yeah, I was in the junior year of college and at our peak had 24 people working for us. I had a partner who was doing the business side and I was on the technical side and brought Kim in to make PC boards. We were making modems at the time and we actually succeeded in making a modem that was capable of going over copper lines at about 15,000 baud, which is even by today's standards an achievement, but that was back in '68 and '69. And so we thought we were really hot and on the way to make our million dollars and we actually pretty well ran out of money because we had this idea that if you could make one prototype that was all you had to do. And we now, of course, everyone knows that's just the beginning of the effort. We ran out of money and we tried to get bought by Motorola and the fellow at Motorola was in deposition and sequestered for the summer so we quietly went out of business and that was the first opportunity we had to really know how it felt to really be in debt, since we owed about, at the peak around 100,000 dollars and we had to settle on a nickel on the dollar to all the people that had invested. We had gone door-to-door selling stock to build the initial equity in the company.

RW: Well how old were you then?

SH: Oh, 18, 19, 20. We used to sell calendars in the Boy Scouts door-to-door and that was a pretty good business so we hit the same houses that bought calendars for selling stock and they paid pretty well. We didn't know that was against the law to sell securities door-to-door without a license, but the side that was really interesting was that we put a little company together. We didn't have a lot of contact with the outside world and we didn't know you couldn't do certain things and I think that was part of the excitement. Well the partner skipped town when the government caught up with us because he'd never paid any taxes for all of these employees or anything, so I got left holding the bag. And it really was a sobering situation, but one none the less that we lived through. Of course, I think that cured my entrepreneurial spirit for a few years.

RW: Was this in high school or college?

SH: College, the junior and senior years of college. So we were working our 70 to 80 hour weeks and going to school at the same time and managing to pass. We took the same courses so that it would be a little easier getting through the classes so that was a lot of fun.

KH: I guess when we got through the Datalink phase and moved on, Shawn had gone out to work out at Martin Marietta for the summer or I guess in the spring. And then inevitably we migrated up to General Instruments. David Chong had been running the MOS operation or the manager up there at TI and he got recruited to help setup some microprocessor design teams right when the 4004 came out and the 8008 from Intel. So we had tried to use these parts and had learned a little bit about them.

SH: Oh, in fact we, Ted Hoff was actually on the road trying to convince people to use his microprocessors and we drove down to Dallas to listen to Ted give his spiel and we asked him whether it would be fast enough to do the data processing for our modem and we all came to the conclusion that it was way too slow so we had to build our own MSI microprocessor or digital signal processor to do the modem. But it got us on the idea that architectures of microprocessors were the way to build extremely powerful computational engines.

KH: At General Instrument we had the opportunity to basically do anything we wanted in terms that their fab was not very good. They were unable to run the silicon gate and the metal peel during the winter because the fab wasn't air-conditioned. But you could go out and run your own wafers and get a taste of what setting up a semiconductor company was like. We also had the opportunity, we put together a 16-bit microprocessor and it continued in that line of just both exploiting what you could do with architectures and seeing how you would get them into production. But we also had the opportunity to meet a number of the people who invented some of the, like the single transistor cell and a number of others who ended up building the UARTS and all of the building blocks that have come to be part of the regular microprocessor landscape.

RW: Well didn't General Instrument at that time was an innovator in many of these areas was it not?

KH: Yes, actually they had what they called the giant process. They couldn't make pin-hole free oxide so they discovered that putting nitride in the oxide would allow them to have a good integrity oxide. And I believe we were running the, I think this was a 10-micron line process at the time and they made Wurlitzer Organ chips and all kinds of specialized chips for the military. In fact, they had a 16-bit microprocessor built to 4-bit slices for the military that would work over the whole military temperature range and nobody knew what to do with it and it included A to D converters at 10-bit level and all this stuff. And the management at General Instrument when the project was over for the military just mothballed all that intellectual property. And we saw that stuff and just went crazy looking at it. It was so phenomenal because they were running 600 nanosecond clock cycle times and this was 1972 or 1971, which that was extremely aggressive at the time.

SH: Intel was only getting 8-bit processors out at about four times slower. So this was such a leap ahead. Of course marketing has a lot to do with your success in this business.

RW: Or a lack thereof.

SH: Or a lack thereof.

RW: Well, you guys look a lot alike, if you don't mind me saying so. Is there a reason for that, or?

KH: We're twins.

RW: Oh.

KH: And I, we've basically followed each other around at the different companies. Learning to work together was a skill we found useful very young in life. It was us against our parents for a long time. And even though we fought a lot it was also very satisfying to be able to accomplish something together. The other thing we found once we got into the business world is that very rarely do you find someone you can actually trust. And many of the businesses if you look over Silicon Valley and otherwise are littered with companies that the founders couldn't trust each other, they couldn't work together.

SH: Yeah, especially when it, when you have to go through the hard times as well as the good times and we had this adage that, you know, two people with the same mind and purpose could always be a single person at any time and that was especially true within the management structures at GI. And then after a year at GI we went West to Xerox. They were starting a microprocessor department there and it must have been about five or six of us all took off at one time to go over to Xerox. That was in '72.

RW: To Xerox where?

SH: El Segundo, it was the old XDS, or the SDS that Xerox bought that was then Xerox Data Systems. They wanted to get into microprocessors and one of the fellows that came with us from GI was a very interesting fellow by the name of Lamar Baker. And he was one of the original patent holders of the one transistor memory cell and he took us under his wings and at least helped me learn how to do something he called reverse engineering. And he explained it all very patiently that you took apart and you took the lid off of it and he showed us how to get the lids off of different kinds of parts, and then you get the microscopes out and you take the pictures, and then you etch it and stain it and take more pictures until you've got an idea of how something is made. And that was a fantastic process. And Xerox being more of a theoretical company than a practical one let us spend a whole year taking apart all of the different microprocessors on the market at that time and reverse engineering them back to schematic. And the final thing that I did as a project was to, we had gotten a pre-production sample of the Intel 8080 and this was just as Kim and I were leaving the company. On the last day I took the part in and shot ten rolls of color film on the Leica that was attached to the lights microscope and then they gave us the exit interview and we went on our way. And so that summer we got a big piece of cardboard from the, a refrigerator came in and made this mosaic of the 8080. It was about 300 or 400 pictures altogether and we pieced it together, traced out all the logic and the transistors and everything and then decided to go to, go up North to Silicon Valley and see if there was anybody up there that wanted to know about that kind of technology. And I went to AMI and they said oh, we're interested, you come on as a consultant, but nobody seemed to be able to take the project seriously. And then I went over to a little company called Advanced Micro Devices and they wanted to, they thought they'd like to get into it because they had just developed an N-channel process and this was '73. And I asked them if they wanted to get into the microprocessor business because I had schematics and logic diagrams to the Intel 8080 and they said yes. Now they hadn't been able to make a part in their fab yet. They had, their claim to fame was copying the Fairchild shift registers and they had managed to make a 1023 and a half bit shift register successfully. One half bit off but nonetheless they fixed that and that was their crowning achievement of their technology, which was a 6-micron process. So in parallel with trying to make a static RAM, they let me have a little group of people and we went off to make an 8080 clone. And one year later we had the 9080 and this was a chip that was yielding about 100 dye per wafer which was about 10 times the yield that Intel was getting on their 8080s and it was four times faster. And we tackled the military market with it and we were selling the first chips for 700 dollars a piece. Unit manufacturing costs was 50 cents. Jerry Sanders was very happy. And they yielded like gangbusters. So the first year of my tenure at AMD was focused on making the 9080 and taking it out to a commercial success. So that was the first step of AMD getting into the microprocessor game and going up against Intel.

RW: Which they're still doing.

SH: We always said we started the microprocessor war between AMD and Intel, which we quite literally did. The interesting thing is that as you, the Intel legal team immediately decided they wanted to tackle AMD with some stop mechanism, but they couldn't figure out how AMD got the Intel microprocessor design. They thought that it came one of the mast designers and so they had an injunction preventing these mast designers from working on the microprocessor project. But little did they know that this came from a totally off the wall source and all of their legal maneuvers really didn't do anything at all in terms of them getting set into the microprocessor business.

KH: Actually what happened was very useful because now there was a second source for the 8080. It was just as good when it came down to it. It was military. And I think a lot of people at those times were very reluctant to use a single source part and that was something that was really a problem to get the microprocessor industry off the ground. You had to be able to show if you committed your boards you could manufacture them regardless of what happened to any one company. And Intel at that time was not as stable, you know, it's not the mega giant that it is today.

RW: Well, in fact you may recall that the first slogan that Intel used was "Intel delivers" and that was because they were sole source on so many products that they needed to have a sure delivery to make it go.

KH: And in fact this was the thing that was causing a lot of problems because they made a kit of parts that were all Intel parts, and if you wanted the microprocessor you had to buy the memory and the interface chips and everything else and this was for the first time allowed the people wanting to use microprocessors to use a ensemble of parts outside of Intel. My second year at AMD I switched over since that mountain had already been conquered I decided to try my hand at EPROMS which was the other product that showed a lot of potential to make a ton of money. And we had reverse engineered the first EPROM, which was a 1702. Unfortunately for the AMD process, it would hold memory for about three weeks at room temperature. Now the spec was 10 years at 85 degrees, so we were quite far away from attaining the reliability goal. And this was where I was able to use some of my knowledge of chemistry and thermodynamics to figure out why the memory was not working and figure the changes that were necessary. But we really, you know, we did so many splits trying to get the right recipe and then finally I was lamenting to my mast designer who came from Intel and I said, I can get everything together except for this one. I've got to know the temperature of the oxide growth. This is the one last thing that I've got to really figure out. And we could do a million different runs, but if we don't get it right we won't have it. And she said well why don't you call Fab 3 over at Intel and ask them what the temperature is. And so I got on the phone and called down to Fab 3 over at Intel and I said what temperature are the ovens at and they said 830 degrees. And I said thank you.

RW: Well so far you've confessed to stock fraud, stealing intellectual property from Xerox, now Intel. This is being taped. I hope you realize that.

SH: Well you know, we talked about some of these things over drinks sometimes with some of the old people that were there at that time and it was quite funny to listen to their side of the story of why they saw AMD all the sudden making some major advances not only in the microprocessor area but in the memory area. And to a large degree they were very loose with their information and they really didn't care a lot about the...how important it was.

KH: You didn't call it, you didn't call it a misappropriation of intellectual property, you called it a drink at the Wagon Wheel.

SH: But the other side of this was that the various technologies where available Intellectual Property didn't really have a lot of meaning back then. And it took these type of actions probably to finally get a definition in place of what constituted theft of intellectual property. We certainly learned a lot in a very short term in terms of the legal implications that go with looking at other people's technology and trying to copy it.

RW: Yet if you look at Europe and Japan they had available to them the same technology, the same brains, the same etc. Except they never made the advancements that we did in the states. That was largely due I think to the cross fertilization from company to company that you didn't have in Europe or in Japan.

SH: Well the old adage was if you wanted to change companies you did not even have to change the parking lot that you parked your car in. You could go from Signetics to AMD and still share the same parking lot, for example.

RW: Well, you worked for Jerry Sanders along the line. Do you have any Jerry Sanders stories?

SH: Well you know I was kind of an outsider and I came from GI and all of Sanders' crew was Fairchild so they certainly enjoyed the development of the products, but in terms of management they were always putting a new boss in and then have to train up a new boss so that wasn't the most satisfying type of solution to the problem. When I started in the microprocessor area I told management that we had to put some major investment into software and also into firmware. And Jerry got up boldly and said our company will never be in the software business and shot down all of the efforts to build a team that could make microcode on an ongoing basis. And that was very disappointing to me and that was one of the reasons I shifted over to the memory area. And when we got the memories finally working, we got them yielding very fine. We were selling the parts for 24 dollars a piece and the unit manufacturing cost was somewhere around 14 or 15 cents and they could make them day and night. The company was going gang busters at that time and it was interesting to see how Jerry was able to parlay that in terms of his political and sales ability. The guy sure knew how to sell parts, that's for sure. But it gave us a real appreciation for the other side of the business. Not just making a technology, but selling it. And this is where Jerry was the undisputed king of the mountain selling the parts.

KH: Well along the way one of the things that we noticed that was sort of a side effect was that an engineer could log on with a silent 700 terminal to a timeshare, spend 50,000 dollars a month, and no one noticed. It was a utility. Now the VP of finance couldn't sign for 50,000 dollars. So we figured if you could, if you had an industry where you had all these engineers who were spending this much money doing software simulation, there was an opportunity there. And the capitalization cost was far less than in the hardware business.

SH: Well in fact in the third year at AMD my challenge was to make the 2708 which was the AK n-channel EPROM, and this was a wicked part because the technology of the way it was fabbed was interval to the amplifier design. And our fab people were good at making some kinds of parts, but they had no idea how to do this type of exotic technology. And needless to say we spent a lot of money trying to perfect that technology and eventually did. But it was curious that again this was an area where Jerry was good at sales, but putting money into engineering was not where he wanted to invest his money. And had he done so he could have, you know, taken them into Number 1 in the EPROM business within another year or two. But there was always this situation where Jerry would make a decision that was going to seal the fate of a whole team and so I think that's even to this day you see that they make a huge advance and they go forward with a tremendous amount of momentum and then they lose the momentum and fall back. This is a characteristic of AMD that's been true for over 25 years now, but it does show a style in terms of management. And we certainly saw a lot of that back in the early days as well.

RW: So what more did you do at AMD or did you leave there after awhile?

SH: Well trying to make the 2708 it, they made transistors. We had 5-micron technology but the spacing between the source and the drain of the memory transistor was, had to be less then 1 micron to get the energetic wholes to go through oxide. And so making a sub-micron transistor in volume in 1975, '76 was quite an accomplishment. A lot of people don't realize that Intel was truly the master of deep submicron and in 1976 and 1977 when they started making these n-channel transistors for their EPROMs. So I tried to design it, but we couldn't make the transistor characteristics between the low gate voltage and the high gate voltage agree in the transistor models, and so we went back to the company that supplied the timeshare computer program and then complained that their models were not good enough. And they said well tell us where the if clause is and we'll change it. And we didn't know that much about software, but we knew they knew zero about the problem. And eventually I switched vendors and they, this was United Computing and they had pretty interesting people over there and they said look, we'll give you, we've got the source code to the Berkeley SPICE program, you get in there and change it as necessary to get your model working and we'll give you a good rate on computer time. And so I learned a little bit about software and got in there and started playing with the program. And as I started realizing, there was a huge need for people that understood the software of the simulator programs but also knew enough to talk to the design engineers to understand their problem. This seemed like something that was exciting to me and then after Jerry delivered his final ultimatum that he would not allow us to go through a very exotic new testing methodology for the EPROMs I finally got sick and tired of training new bosses and decided in '78 that I'd go off on my own. And the timeshare company said well we'll let you use our computer and you can have all the rights to the software development as long as you let us timeshare the program to customers because they just wanted to sell computer time. Now they had just gotten a brand new computer and they had no one using it and it was called a CRAY 1, and this is the first commercial CRAY 1 in existence. And it was sitting out there in Kansas City, no-oping itself to death because nobody was using it. So we had an old 110 baud modem with a teletype machine. And we could compile and test and do all of the exercising of the software and the entire cycle between making a change to the program and compiling it and testing it and being able to make the changes to the program again was about somewhere around 18 to 25 seconds. And they had the fastest compiler and the fastest machine, so we learned something about software development and that was the number of turns per day that you can do with your software and this was the same with your circuit design. The more times you could iterate on the design per day determined how good the design was going to be and this was true for our microprocessor design as well as the memory design. So we took that idea and we'd work our 80 hour weeks and some days we'd make 4, 5, 600 iterations to the program and that, we realized, was our secret sauce. That everybody else, no matter how bright they were, if they didn't have the number of iterations to understand the program it wasn't going to go anywhere fast.

KH: I think one of the things that, if you look at the cycle time we had back then and if you look at the "software engineers" have proposed, many times you're lucky to get two or three turns a day with all of the heavy overhead of software engineering applied to what's really a much more mundane problem to solve. So we saw also that being able to take the CRAY 1 via timeshare and test out a lot of the more advanced ideas, we could discover new ways of doing the programming, which was a vector programming techniques no one else was using. So we evolved our own circuit simulator, called it Meta circuit and it was about 6 to 10 times faster then SPICE. Our dad would come out and help us set up the books and he offered a little bit of advice to us. He said, well what do people when they do circuit simulation, what program do they use, and I said, well, SPICE from Berkeley. They sent out 5,000 tapes. He said well, whatever you do, call your program SPICE because it has name recognition. So inevitably we modified our program back into SPICE and called it H-SPICE.

SH: There is a funny anecdote about this first product because we spent about two years perfecting this new kind of circuit simulator that was vector based. Now the entire customer base of the number of companies that design circuit simulators that had vector computers, that entire, the entire world market for that was one company. And they already had a circuit simulator. And so we at that time learned another marketing principle and that was that you had to have a customer base for your product. And that sealed the fate of our Meta circuit program and that was why we went back to SPICE and started building a program that could be ported from the CRAY 1 to a VAXs which was becoming the popular mini computer of all of the engineering departments. And within a year we had a working version of our product on the VAXes and the IBM mainframes.

RW: How was it different than UC SPICE?

SH: Well the, one of the things we said back in the early days was that we could give the program away. It was the consultation and helping the customers use the program that was the real value added. Now we did make a tremendous number of changes to stabilize the program so that it would work reliably and give consistent results, which the university product tended to be pretty buggy. But the real focus was being a circuit designer and having a circuit designer call up with a problem and being able to diagnose what was wrong with his circuit as well as what was wrong with our simulator. Because sometimes he had a nonsense circuit that the simulator wouldn't be able to solve anyway. So we had to have enough understanding of the design techniques and the kind of circuits to get in and diagnose the problem whether the problem is a user problem or our problem. And that was one of the things that I think we got the most enjoyment out of was, what's now called hotline service. And people would call us up with their problems and we'd sit and chat with them for a while and get to understand what they were doing, what the nature of the problem was and then eventually we'd get it fixed. And because we had this idea of very fast turn around time, we could typically solve the problem and have them fixed the next day. So we delivered good service and we charged 24 percent of the software cost every year for service maintenance contract which was considered high by anyone's standards but because we were frequently working late at night, and you know, the engineers were also working late at night and they were, you know three weeks late on their projects and knew they could call us up and get someone to help us out, help them out. And so they gladly paid for that kind of service.

RW: That's great. Going back to EPROMs, at Intel we discovered that the, if the devices were in, under fluorescent lights that they would erase in about a year and in sunlight they would erase in about a week. And at that time we had hundreds of thousands in service all around the world and so we came up with a little sticker to put on the lid, on the transparent lid to stop that, but we didn't want to admit that these things were erasing because in fact most of them were in boxes with doors closed and they didn't in fact erase. But, so on the little sticker we put on there the code number and the revision and all this neat information so they would have it right there, and then we gave all these stickers away. So I don't, did AMD ever figure that out? We were doing that?

SH: No, but we did figure some ways to do some very rapid testing that would get the cost down and then they tended to, we eventually got very high reliability and learned to play with nitrating some of the passivation layers that ameliorated some of that effect. And again this was an area where AMD's upper management was really focused on selling product rather than researching the cause and effects. So a lot of these things didn't get the attention they needed.

RW: Well, so when was Meta Soft actually founded?

SH: Well we started in '78 as the Hailey Company and we went and tried to incorporate that after a couple of years and there was a travel agency in Palo Alto called the Hailey Company or Hailey something so they said that's not going to be a good name. So we went back and thought well we wanted to have a name of the company that would be interesting, that would be, that you could remember but wouldn't tell anybody exactly what you did because we didn't know whether we were going to be selling software or hardware or, you know, getting into the potato chip business the next year. So we had, I had played with some Meta compilers back at Xerox, and so the name Meta, and of course that in chemistry is a, is one of the designations of where the molecules are ordered in a molecule. We looked that up and it said over and above as the Greek translation of the world Meta, and so Meta Software sounded like over and above software. That made sense to us because we were really focusing on the end application of the software and not just being a software company in and of itself. So we tried the name Meta Software and everybody seemed to like it and remember it. But of course it was interesting, after Microsoft became very much the name everyone thinks of, people would say Meta Soft instead of Meta Software because it sounded a lot like Microsoft so we, I guess got a lot of extra name recognition because it was similar to a more famous name.

KH: I think also, all the engineers use H-SPICE so I could always meet people on airplanes who'd used H-SPICE. It was very difficult to get the name recognition of the company aligned with the name recognition of the product. Engineers are that way.

SH: So we would always say we were the H-SPICE Company and they'd say oh yeah, H-SPICE, I know who you are. And they'd say your company name is what now?

RW: Your product would be purchased by relatively few people in absolute terms and, of course, even within the semiconductor company how many people would really be in charge of the circuit simulation end of things. It'd be just a handful.

SH: Well it's interesting because in most of the companies in Silicon Valley the circuit design engineers were held in such high esteem that the tools, the software and hardware tools that they required to do their work set the standard. So management would let the engineer decide the tool that should be used. This was especially true at AMD although at the companies like Motorola and Intel they built their own. So the engineers were basically forbidden to use outside tools for many, many years. But all the startups after that, especially sons and daughters of AMD and all these other spin-off companies, the engineers that had used H-SPICE at AMD or at any of the original companies that we sold to could move to a new company and say oh I need to have my tool come with me. And so they would be the biggest broadcaster of that. And we even in our marketing ideas said, well, if we sell our product to certain companies they will force a tremendous leverage of use of our product. In fact, that's why we targeted a company called LSI Logic because they were a company that sold a software service and built products for other chip designers. So if I sold one copy to LSI and LSI had 25 companies using the services and the 25 companies using the service would ask LSI what circuit simulator they used to validate with, well they would have to buy the software as well because the standard response at LSI would be, well we trust H-SPICE and you can certainly use any other circuit simulator, but we'll resimulate it in H-SPICE just to make sure. And that gave a cache of stability and authority and all these other companies that were less sophisticated would use what worked. And so we had a built in marketing through our original customer base.

RW: Well at LSI, we had our own software development, and our own software sales, but H-SPICE was one of the few products that were specified by the engineering group and our sophisticated customers could request the H-SPICE parameters of various cells so they could do their own analysis perhaps using them in ways that we had not anticipated. And, but it was, that and Kelma really were the only outside software that was permissible. Everything else our CAD people killed.

KH: One of the reasons I think Meta Software had such a long run, we were 18 years profitable every year, year after year, was simply that we combined the software, the knowledge of circuit design and the modeling. When we first started selling into companies we would make sure we had models that worked and we'd work with their fab in order to get good model parameters for that fab. Many times our business side took the back seat to getting good results that the engineers could trust. When we finally went into AMD, the business argument was it was a 9 day payback over what they were using without side timeshare services. They started out with, I think buying something like four million dollars of outside timeshare a year and they were supposed to go up to eight, and we started out trying to sell our product for 30,000 dollars, which is sort of a funny story when we tried to come up with what is the price of software. Well, the reel the software comes on is only six bucks so there's not much value there. You try to impute the value, we start at 15,000 dollars and we didn't sell any copies of H-SPICE. We moved it up to 20, didn't sell any copies. Moved it up to 25 and sold one copy. And then we set it at 30,000 dollars and we sold, you know, hundreds of copies. So there's a point at which people have some notion of what something should cost. And also we came to understand what it took to support and improve the product. One of the problems that the software industry has is that they typically don't reinvest in their product. So you'll see a company who comes out with a product line, they'll take that product line and they'll run it for about four or five years and then they'll die, they'll go out business. So being able to have a software product that's in continuous redevelopment year after year was one of the things that made I think Meta Software a little different than some of the other companies. Also, the tight relationship we had with the foundries and with the people developing models. We had our own modeling capability so we could have wafers fabbed in some foundry with our test pattern on it and then take those wafers back and develop models for that process. Then that was pretty bullet proof and the foundries appreciated it. And the engineers developed a tremendous trust. We've had circuit designers who every company they went to they would specify they want to buy a copy of H-SPICE and engineers would stay at a company maybe two or three years. So we've had some engineers who've bought literally hundreds of copies of H-SPICE for the companies they've worked for.

RW: So what were your sales like and who were the sales people? Were you people...

SH: Well we basically sold it ourselves. We sold it by telephone and so, because we didn't want to, we had to answer the hotline at the same time so we would just call people up, call the engineers up, ask them, you know, what they were using and talk to them because as ex-circuit designers ourselves we were extremely comfortable talking to the engineer. And I remember as a circuit designer myself, when a salesmen would come in and want to talk to the engineers themselves, I mentally had in my mind this lever and I'd give the salesmen about fifteen seconds as his lead in and in my mind if he didn't say the right things, I'd pull that lever and I'd mentally saw this trap door open and that salesman would go right out the trap door. I think the point that as a circuit design engineer, whenever that salesmen would come in, you'd give him about ten to fifteen seconds to say something and if he wasn't technical, in your mind you had that lever and you'd pull the trap door and your mind just went to some other place. And the salesmen could talk for an hour and you weren't going to register anything and you'd find a way to get rid of the guy as quickly as possible. And some engineers are more tactful than others but generally the salesman ends up feeling pretty well beat up if he's not able to talk the language. And with our product I think our H-SPICE product was about as technical as you could get being sophisticated software for circuit designers. And so having that background and feeling comfortable and not like I was going to be eaten by the tigers, we could go chat with the circuit designers or the engineering management and talk about their problems. And we loved to talk about it because we'd been handling all their hotline for many years and, you know, we'd say by the way we know another customer that has this type of problem and, you know, the way they solved it allowed them to really make a lot of money and maybe we could help you as well. And so we didn't know how to translate that idea of being comfortable in front of the customer and being able to chat with them and talk about their problem. So Kim and I ended up doing most of the sales.

KH: In the early days, and I think as the company evolved and we were able to bring on most of our sales people had electrical engineering degrees and were maybe not as comfortable as we were, but at least they didn't get beat up as severely as a regular high powered salesmen. I think what we have learned though over the years is that salesman are a different type of animal and you know you can hire good salesman and they can go sell ice to the Eskimos. But to us there was a certain integrity and also the feedback was very essential from the engineers themselves.

SH: Yeah we had a policy and that was that we refused to accept the purchase order until the engineers had actually used the program and felt that that was the best solution to their problem. In fact we had one customer, Samsung, and we said no, we want a purchase order with no conditions on it and we won't accept your purchase order until your engineers are completely satisfied. Well they went on using the program for two years and then finally we had a time bomb that went off and we, they called us up and said you've got to turn it back on immediately, and I said well are you completely, absolutely sure that your technically satisfied, and they said yeah, yeah, yeah, get it going. And I said well what's the rush, and they said well I've got 40 engineers who've been using this full time for development for the last year and they're sitting on their thumbs. We need it turned on immediately. And I said well get us a PO and we had a PO come in about 30 minutes later and that was the kind of methodology we found most effective. We didn't have the business orientation to say you've got to have so many sales this month because we knew we would get the customers sooner or later. Many, many companies had their in-house capabilities, Intel for one. Many, many companies had their in-house SPICE programs. So until something happened, a lot of those companies we couldn't get in and sell. I know there was one company, Rockwell, and there was a gentleman there that ran the CAD department and he personally had developed the Rockwell circuit simulator and we went and called on these people every year, must have been for about five years. And I remember remarking to you that we're not going to get into this company until this guy dies. And then it took another two years and the guy finally died and we got the order within a couple months after that. So literally, you have to wait a long time for some of these things to happen. And with today's high-powered salespeople I don't think they really understand there's a natural progression that you can beat your head against the wall but nothing's going to happen until you finally have that critical event.

RW: Well you were privately held, so you didn't have to meet any sort of growth criteria set by investors.

SH: In fact, Kim and I were the two partners essentially. We were a subchapter S corporation until the night we went public. And so Kim and I were the only shareholders in the company as a private company. And the interesting thing and I think it was more because we didn't have any real competition for so many years is that over the 18 years we had a compound annual growth rate in excess of 25-30 percent every year for 18 years. And if you look at any other company that's a pretty good record, although, we started with a very small...

RW: Small base.

SH: ...initial capitalization. We kept it growing every year and I think it grew a lot because we weren't focused on making the numbers but more on making a good solid customer base and really spending the time with our end users and making sure they're absolutely happy.

KH: The circuit designers are the real heroes I think for the entire electronics revolution and they're probably the most under appreciated. They're not the people who take the chip out to the market; they're not the people who get that big 10 billion dollar order. They just silently grind away in the background the actual circuit. And there's the creativity that goes with that but it's also the knowledge, the intuitive understanding of the electronics and the physics and the mathematics behind it. So I think those are the type of people that from our standpoint, they were are customers and we couldn't of asked for a nicer bunch of people to have as customers. They're just people we've developed relationships over the years, and just the nicest people you could imagine.

RW: You've been so close to the industry since it was very small, and since they were working designs had a few thousand transistors and now they have a few million. What changes have you seen during that time?

SH: It's really I think, for our industry almost no changes. I called on Mosel in Taiwan earlier this year, talked to the designer who was designing their next generation DRAMS and they were still using basically the same model that they had been using fifteen years ago. The amount of change is been in how you assembled the design. You take the latest microprocessors and maybe they're 64-bits internally instead of 8 or 16 and they have a lot of RAM on them. But the real difference has been that architecturally they're much more complicated. For the circuit designer, the task is almost identical to how you get started back 20 years ago. And I think that's why the longevity of the SPICE in general, it's a process that isn't going to go away. And when you look at the next generation parts with the deep submicron revolution, it only makes the SPICE business bigger.

KH: Yes we've been, I would normally go out and do the training, the technical training for the company and especially over in Japan and Korea, I've been training these guys for the last 18 years literally, so we've had almost one or two generations of management where the engineers have made it into management and these were they guys that I trained up 5, 10, 15 years ago. Of course for the Pacific Rim this is the memory business and our business went west basically in the early '80's. So we had a significant, probably well over 50 percent of our total business was Japan, Korea, Taiwan. And we've made a lot of good friends over there and have had a lot of very interesting times. I remember when we first decided we wanted to have representation in Japan, I said well it's time for us to get a Japanese distributor so I got on the airplane and flew over to Tokyo and I got a room in the new Otani Hotel, got up to my room and pulled the, they had a yellow pages out and in each room Japanese yellow pages and it was in English! So I started cold calling companies out of this copy of the Japanese yellow pages and of course I'd heard all these things that you had to make appointments at least two or three months in advance to be able to get to see the Japanese. They were so formal and everything. But I think they were so astonished to hear someone just calling them up and they said well where are you and I said well I'm at the new Otani, can I see you this afternoon? They were flabbergasted and they said, okay. I guess the manager over at Nihan Digital said I think I need to take care of you because this is very untraditional way of working business with the Japanese. And so he became sort of my mentor. And I said well, you know, I really need to get someone out here to do this business but they've got to be so technical because the traditional sales in Japan is with a trading company and I knew they wouldn't be able to technically solve the engineers problems. And so I said I need a company, not a trading company but one that could be a technical, viable force in Japan, and the guys says, I'll find you a distributor. And about 3 months later he called me back up and said I have your distributor. I've found your distributor for you. And he didn't even tell me who it was. So I said this is really interesting, he says, come to Japan and I'll let you meet them. And over that time I had learned about traditional Japanese marriages and I realized that this was an arranged marriage. And I went to the office at Nihan Digital and went into a special room and there were these 4 old guys, all of them were over 60 years old and one of them got up and in broken English says, 'we're the old boys,' and the company that they had selected for us was a company called NTT or Nippon Telephone and Telegraph. Just like AT and T they had a labs, the NTT labs. And as a spin-off of the labs they had what they called the NTT technology transfer business where they took the inventions out of the labs and put them into Japanese industry. And they wanted to, they felt that they had the technical expertise to be able to solve our problems.

KH: They weren't IBM. The selection was not to have another computer company represent H-SPICE because the exec guy didn't want us to fall into Sun or Apollo or any of these other emerging companies so he hooked us up with the telephone company.

SH: And the fellow, one by one they each got up and told us what the guy had done and this one fellow stood up and says, I invented the first 1K dynamic RAM in Japan. And the next one sat up and says I developed the first sync amplifier for core, magnetic core memory. On and on, each one of these guys had been a distinguished PhD level fellow within the labs. And we found out that once they hit 55 they were given the opportunity to join a start up within the company. And so what they did to make new companies back at that time was to take half old guys that were approaching retirement and half college graduates, put them together and make a new little company. And this was the NTT technology transfer company. And they had a quaint little marketing and sales outfit. They had a little pamphlet that they sent out every other month and at the top of it said "to the friends of NTT technology transfer society." And it went to 400 people and they were all VP level at the major electronics companies. And...

KH: NTT had licensed all of the early DRAM designs into Japan...

SH: Like Hitachi and Toshiba and all these companies got their intellectual property from NTT.

KH: And that's how Japan got into the business. Of course they went out and bought all new equipment so they could build these designs and put the United States out of the DRAM business, but that was a good entree. It wasn't through the front door it was through the back door.

RW: But you were happy with what you got it sounds like.

SH: We were. They, it was slow getting started. These guys, you know, when the fellow from Nihan Digital called me up he says look, these guys really know technology, that's the good point. On the bad side, they don't speak English, they don't know anything about sales or marketing. Which was true. But what we found was that allowed us over the years to establish very deep and long personal relationships with key people within the companies such as Toshiba, Hitachi, Sharp, and that's resulted in over the years just a very strong allegiance to our product. And nowadays I think that just about every major company in Japan and Korea is still a customer of our product. In the same, actually it occurred in Korea as well. Once we had convinced Sharp after two years of evaluation to use our product, all of the other four major companies in Korea bought at exactly the same time. And the reason was that they all went to school, the CAD managers all went to school together. They were all in the graduating class and whatever Samsung did everybody copied because the government mandated in tools and equipment that all these companies would use the same vendors so that they could put the combined power of all the companies in Korea behind one vendor.

KH: I was at the design automation conference last year and the CAD manager from Samsung came up to me and said you know, ten years ago we got into the DRAM business, our company had never done anything particularly aggressive in electronics, we're a ship building company. And he said, we started working with Meta Software and H-SPICE and we could not have built our first, we could not have started our DRAM business and built it to what it is today without H-SPICE. And that, last year Samsung was number 1 in the world in DRAMs. Of course, I said can I quote you on that, and he said yes. But the point was that we helped a lot of these, say, emerging companies get their feet on the ground and literally Meta Software and TMA who had all the modeling software for the fab had a tremendous loyalty in the emerging growth companies in Korea and now in Taiwan. But those types of relationships are the ones that we felt I know personally, most satisfying.

RW: Well what happened with Europe through all of this? What, they at one time were a major semiconductor manufacturer in a number of them in Europe and now they're pretty much gone aren't they? Phillips and...

KH: Oh Phillips is very big and Phillips is in my estimation one of the more innovative companies. They don't get the press over here. Their TVs obviously aren't sold into our market because of the differences between the C-CAM PAL and NTSC. On the other hand, I think their politics have gotten in the way of good sound business. And literally have not invested in the same manner which the Taiwanese, Japanese, and Koreans have.

SH: There's also, I think a level of aggressiveness that's been lacking. It may be because of the government structure or whatever but it's frankly, these guys don't work late. They take off their holidays. They are very, very laid back when you go over there and you just don't see them really busting their butt getting product out the door. And you go to anywhere in the United States or in the Pacific Rim and people hustle and that's missing over there.

RW: Well look where they've gone to in computers. From some place to no place. Well at some point in time you decided to sell the company.

SH: Well let me take you back a few years that's an interesting one. For me, I was in a position to guide the company as its president. My concern though was that I was fundamentally an engineer. I enjoyed talking with customers, but a businessman? Not really. So I ended up taking a lot of courses to learn more about business and as we went through the threshold, I think it was around the 8 to 10 million dollar mark in annual sales that we finally started realizing that we had to put a fundamental structure together. We were profitable, we were making probably, when we were making 10 million dollars in sales our profit before taxes was around 4 million or 40 percent, which is reasonably good. And so the earnings were quite fantastic for a little company. I realized that we needed to have some backups for Kim and myself and it was the old adage, what happens if you get run over by a cement truck on the way to work? What's going to happen to the company? And so I started at that time, this was about, oh seven, 1990 or thereabouts I said we need to put an executive team together and get some real people that are veterans in the business or otherwise there's no way we could get out of it. And I was working you know 70-80 hour weeks and my wife said look, you've got to start getting a little bit of time with the family or by the time you finish doing your job there's not going to be any family left. And so that was the point where I really saw a need for putting a team together. And we first, our first executive hire was a VP of sales. And we had a recruiter come in and help us and he got three people that he felt were good candidates and it's funny. The first one had been, all three of these guys had been salesman at Daisy, which had eclipsed at that point in time and was no longer a viable entity. And the first guy came in, gold chain, looking good, nice sun tan, and he says I took it from nothing to that maximum point where we were really making tremendous money and I took my options and I got out and made a ton of money. And then the next one came in and said I took it to it's zenith and we really made, you know, it was at the peak and then I cashed out and he also had a good tan as well. The last guy came in from the Midwest and he said, yeah I was heading up sales when we were chapter 11 and I was having to make sure that we met that quarterly goal even when the company was literally almost out of business and so we really had to scrape hard. And Kim and I being from the Midwest and certainly could really, really understand this guy and where he was coming from, it was an automatic. We said we want that last guy on the team. And that was Al Lepinsky and he was our VP of sales up to, right before we went public. And then we learned about options because you've got to incentivize people and we had been going on for about 12 years at that time without incentivizing our employees. No stock, we just gave good salaries. We figured that was good enough. But these guys wanted stock and so we put an option plan together and being a sub-s corporation we had some real severe problems in doing that so we invented a hybrid that was an option only if we went public, otherwise we would make a market for the stock ourselves and buy it back if they ever left the company. We put our team together but that set the clock a'ticking because these guys wanted to see the company go public. And I think that more than any one thing set our course. And within four years we went public and all those options became good and everybody did quite well by it.

KH: I think one of the problems we always had to deal with in any sort of a high growth industry is that you've got to get the talent. You've got talent, you've got timing are about the 2 variables that you have a little bit of control over. And of course you have to have the money to be, and the where with all to do it. So as people become available you can grab them many times and if you've got a plan that's exciting and can articulate it in a manner which isn't too confusing, then they'll jump on for the ride. And then instead of the business being what we called the Dolby labs type of model where you just grind on and on and on and it's not the superstar model, you have to start adding in how do you get this long-term growth model to take off and be attractive to these people. So two years ago we were able to generate enough interest in the company to take it public.

SH: That's a funny story because the, when you get ready to take a company public you have to go on the horse and pony shows and go out to meet the financial analysts and everything so our, the first step is the American Electronics Association Conference in Monterey which is where everybody goes to test the water. And they have a five minute limit on your presentation and I was about midway in the pack which put me on the second day all of the venture capitalists and the investment bankers were there and everybody had maxed out on their five minutes with these pretty boring long slides about their technology. And my angle was, I said my technology they're not going to ever understand it, it is rocket science, so let's not even tell them about the technology. And so I just said, went up to the guys and said let me show you our growth rate over the past ten years in sales and earnings. And I put the slides up for the sales number and it was one of those exponential curves and the earnings laid right on 24 percent compound annual growth for 10 years and I said thank you very much. It was approximately what, one min and 30 seconds and I got a standing ovation because these guys were so numb from all these long product descriptions, they didn't even care about the product. And that was my first introduction to understanding investment bankers, what motivates them and what they're looking for. And shortly after that we lined up three different investment banks, Wessels, Needham and Soundview and then in, what was it, November of '86, we went, no October, we went on the road show and we,

KH: '95.


SH: ...'95. We were 9 times over subscribed. We've put about 20 million dollars worth of stock on the sales table and the investment bankers helped us, we went on the grueling 2 week tour and then we went public. And that was pretty much a real high point for us, really seeing the translation of all that theoretical wealth into something that had a little more tangible value. It was actually quite shocking, the tangible value.

KH: But it also set a clock ticking, which we also knew with the next generation of processes, and ultimately, circuit simulation was subordinated to place and route because of the electrical domination of wire over transistor. When we started transistors had 80-90 percent of the delay for a circuit. The wiring you could almost ignore. Rule of thumb at 10 percent. That shifted over the years and as we saw a quarter micron and below we could see 25 percent of the delay being in the transistor switching and then the majority on loading. And to tell you the truth the problem became almost over powered by placement and routing. And of course place and route was in trouble. They had to get more of the circuit analysis embedded into their tools. So we saw only two candidates long term and that was Cadence and Avanti. And Cadence I think was a little bit more arrogant in those days. They certainly felt they had dominated the marketplace and didn't have to worry about any changes in the technologies coming if all they felt they had their place and route was number one. And even though they were a little weak in circuit simulation that's where the company's origin was. Jim Solomon was a circuit designer. He had guided the company and I think though he was no longer as active in the company, that legacy was sort of a built-in that they knew how circuits worked. They hired a bunch of people from Harris which were also excellent circuit oriented people.

SH: But you see that was the technical side in what we had to look at was the market dynamic. I mean we can go technical, it doesn't tell you what's going to happen to the business. We looked at all the possible places where our technology, our products would be a perfect add on match. We went over and to talked to the guys at Synopsis and they didn't understand the technology at all. Cadence had a competing technology. Avanti was the only other guy out there and so we decided that is one we should get with. Now this was right as Avanti started their biggest legal problems, and the, I remember the day after their stock crashed from something like 55 dollars to something like what, 10 dollars? And it was when the police came and raided the building and had everybody with their hands on the wall and they frisked them down looking for intellectual property and took over the computer systems and backed up all the disks and everything. And that day was also odd, we had a power failure in San Jose so all the telephones were down. And so everyone that wanted to call Avanti to see whether the rumor was true or not got no answer. And so the investment bankers freaked out and all the sellers sold and so the next day I gave a call to Jerry Hsu who was the CEO of Avanti and I said well, I don't think that you're going to be seeing too many of your so called friends around today, I figured I'd give you a call and tell you that I really understand where you're coming from and we should get together and chat. And I think that so touched Jerry Hsu because in fact nobody wanted to talk to the guy being implicated in the theft of intellectual property is like a, being treated as almost a common criminal. Everybody that had previously been talking to him said I don't think we should be talking anymore right now. But that set the interface point and within about six months of that telephone conversation we had had few meetings. We enjoyed each other's company. We both saw the market requirements regardless of whether this company was having legal problems or not. They were the number three company in the business and the only fit that I could see for our company. And so that was the beginning of our effort to try and get married. And it also solved another problem for us and that is as the dominant share holders in the company, there was no way we could get any liquidity without having our stock go down severely because everybody looks at insider trading and when you're the significant owner of the company and you start selling, that is a very, very bad signal. But if we're merely a subsidiary of a larger company that gave us a long-term way of eventually moving out of the business.

RW: So you then did sell?

SH: We, yes we did. I was actually the, I had talked to our investment bankers and they were just scared stiff about all the legal things going wrong with this company that we had chosen to get married to. And they said no way, you know, there was no way they could feel comfortable in dealing with the legal problems of this magnitude. And so the summer came and as the summer was going through, it was August time frame and actually I was out here on Kauai taking a little vacation break and the, I got a call one day. It was about 8:00 in California, 8 in the morning California time, so it was 5:00 in the morning here and it was the investment banker calling and saying that even though we had met our quarterly earnings they felt that our stock price was not sustainable because we didn't have a good story to tell about why this one product company that sold H-SPICE would continue to grow at 30 percent per year. You could tell them that, but there's no believability there. And they said they didn't think they could hold the price up more than a couple more weeks. And so I flew home two days later. Monday morning I went to my office and called the investment banker and told the investment banker that I was going to go sell the company that day. And I said look, I'm going to go over at noon and you can either come with me or I'll go alone. And the investment banker, this was Mike Ogborn, he looked at me and said you're serious aren't you? And I said yeah. I said well, I said look, go in the other room, go talk to your boss, and see if you all can't come in with some kind of a solution.

KH: And we had to get a fairness opinion that this was a reasonable merge. And they were really scared. Of course they had been taken Arksas public and they knew there were some legal problems over there. And of course we had our similar difficulty with getting consensus with our board of directors. But you know, when you look at it and you understand the best you can about what's going on within a company and what the legal ramifications are and you've been in business long enough to know that a lot of the problems are not nearly as serious as they may appear. You know, you get this, a lot of people have gone through all this of trying to make everything look very clean and squeaky clean on the surface. When they look at well is there a problem in the company? Is there misappropriation of code? When an engineer goes from one company to another company as often as they do in Silicon Valley, we call it cross-fertilization; sometimes the police have another word for it.

RW: Theft.

SH: Theft!

KH: But I don't know of anyone who left a company who didn't take something with them, either tangibly or intangibly. And that's part of what it took to make the whole industry grow. Otherwise, companies get sedentary, they rest on their laurels, and the customer gets pretty dead service at the end of the day.

SH: Well this brought up the point we talked about in terms of the intellectual property and we summed it up real simply, was this company dealing with engineers that had to have tools of the trade? If I'm a carpenter I've got to have my tool kit I've got to take with me to work and I can take it from one job to the next. Or was it a trade secret? And so all of these engineers had actually done this kind of work before in college for their Masters or their PhD thesis and so for a company to say you stole our trade secret, the guy's coming back and saying well it's just a tool of the trade. I've got to be able to do this; this is what I learned in college. So that made it doubly hard for us to figure out what was right, whether there was in fact a problem or not. But we did know one thing for sure. Cadence definitely thought it was a theft and the district attorney was reasonably convinced, but at the same time I actually went to Joe Costello at the time because I still wanted to give Cadence one last chance at buying us and Joe said don't go with them, you know, they're liars and thieves and all this stuff. And I said well, I'm not hearing any words from any other suitor and I said also I have at this point what I call a Bhopal mentality, and this was in reference to Union Carbide. They're stock was depressed severely after the bad chemical event in India that killed so many people but, after it was all said and done the company actually survived quite nicely and kept on going, and particularly legal situations have a way of being resolved business wise before the worst happens. And that was our thinking at the time, that the, it looked a lot worse then it really was.

RW: Yeah, well now you've had what a year to think about it? It was the right decision, right?

SH: I still think it was the right decision even though they have continued to have difficulties. I think they're working them out and hopefully, you know, the companies will find a business solution.

KH: When you look at the problem in general that Meta Software was facing, we were no longer facing particularly technical problems in as much as there was a bigger business problem looming on the horizon. We had gotten up to 25 million dollars and didn't have a product line to go afford the type of salesman required. And we looked out on the horizon there and there's Jerry Hsu who's probably one of the most innovative salespeople our industry has seen. And I think that to his credit, our industry is evolving very fast in terms of how people buy software. They don't buy it, they essentially are leasing because the software and the industry's changing too fast. So we have to keep in mind that the problems that the ECAD business are facing are both technical and a business. And the business has got to take front seat right now because you're looking at these companies surviving and with the consolidation in our industry going on it's imperative to go ahead and help push that forward. I think on the other hand though is every time you get the consolidation it gives the opportunity for more of these little start ups, the innovation, the creative juices get going again. So I think that's where we actually headed and we did the right thing.

RW: As we wind this thing up, would you do it again and if so, what did you, what do you feel like that you did that was different?

SH: You know what I liked, I really enjoyed what I was doing and really didn't care about the money. And as long as I was really enjoying what I was doing, that made all the sense to me in the world. When you start looking at what am I getting out of this or what am I making out of this or something like that, it's not fun in my definition at least. And that's still going to be my guiding principle and if I find something that just excites me and is a lot of fun, yeah, I would do something. And we're kind of looking around and saying well, we're still not totally over the hill. Maybe there's something fun to get into that would be exciting and get us into some endeavor that we could really enjoy. But I think it's got to be for the fun of it.

RW: What do you think Kim?

KH: Well we got into electronics through more or less as a hobby. And I always think of it as the world's best hobby because it supported me for all these years. On the other hand, there's an intensity that the Bay area has that allows anyone who's being reasonably successful to take it to its fullest potential. And I think when Shawn and I moved up to San Jose and saw what was going on, it's very easy to create a very exciting company. And there's so many times where even though we may have been a little bit, you know, solving a very tough problem and not having a lot of fun at that piece that the challenge just was a constant push to keep going. And then your customers pull you forward. And I think if nothing else, staying in contact with those customers or developing new customers, that's the fun part. You know you're doing something for someone else. I remember getting on a plane a couple years back and seeing, talking to the guy who sat next to me and he was customer and he had read our manuals that had been written over the years by a lot of effort on our part and also seeing that he depended on this and I think that's what for the most part makes it worthwhile. I think as far as where to go next? In the Bay Area it's so much opportunity. At this point in our lives we get a really nice break, take a little time off, see what's going on, see where we want to take our lives, spend a lot of time with our families, having this Christmas in Kauai has been so much fun. It's a great place.

SH: The other part to me at least is that there is I think a significant change going on with the development of the Internet. I think that coordination and development potentially of a company that is web based is no longer fiction but actually feasible. And one of the areas that we're seeing is that I'd like to have a business that could be portable. And I was so thrilled this summer when I got my first laptop that linked into my cell phone and realized that I could be on the beach on the Internet downloading and then I had five hours of time, battery life on the system. So I think this is now ushering a whole new era of a globalization at the individual level that we may be able to take advantage of and that to me is a really exciting concept moving forward that perhaps the whole design business is potentially ready for another way of looking at putting businesses together.

RW: Well let's see what the wives think of this. So Kim, would you relinquish your seat there for Jan? So who's this?

SH: This is Jan, my wife and she's been through the thick and thin of most of Meta and certainly has seen the ups and the downs.

RW: Well now you're here in Kauai and you're getting some of the ups presumably, some of the good times.

JH: Yeah, it's been a long road. Shawn and I got together in the early days of his company. I think one of the things that really makes it a quintessential Silicon Valley story that Shawn didn't mention is that all these early days of activity when he mentioned the engineers who'd call him up and talk to him anytime. It was taking place in our kitchen and our two back bedrooms. So this is the reason that when the East Coast woke up they could call us and get us and when the guys in Japan were up late or wherever they could still call and get Shawn anytime of the night or day because we were basically sitting around the kitchen table and had the old IBM electric typewriter typing out the manuals and I think if a lot of these companies in the early years knew where they were posting those checks to they might have had second thoughts about doing it.

RW: Well we didn't talk about this but the story that is often neglected is the families, the divorces, the stress that one goes through, it is tremendous amount.

JH: Right. Well in the early days, you know, before we had any children, all that was okay. And then our firstborn Neal came along and that required one of the bedrooms so things got a little cramped and it was really tempting to say, oh here Shawn, hold the baby while I do this and then the phone would ring and Shawn would say well you have to take the baby. So eventually that led to a move to Kim's house in the garage. And this always astounded the deck VAX's repair man and stuff when they were called to an address and here it was the garage in the back of a house in Los Gatos. They were not quite sure they had ever arrived at the right place.

RW: And it has a half million-dollar VAX that is not stolen.

JH: And a few guys working away back there.

SH: And remember the engineer from France that came and got the wrong address and it was a bunch of chicken. It was a house that had the chickens in the back instead of Meta Software. But we finally got him straightened out.

JH: Yeah those were funny. Well you mentioned the stress. I mean those things, you know, working 60-80 hour weeks is pretty much the norm in Silicon Valley and anyone who's there does that. And I always kept pushing Shawn to try to find a middle ground and Shawn to this day maintains that there is no middle ground. You're all the way on or you're all the way off in Silicon Valley, at least, for his mode of operation there was no in between. And so I was talking to your wife earlier, I was saying I think it is hard to find a balance in Silicon Valley and we're probably not balanced here. We had to go to the complete other extreme and be away from it all for a while to start to figure out where that balance point is. But I do know certainly many families where, you know, the husband's gone through the whole stress of taking a company public and then they turn around and do it again. And it's really not the money, it's the ego thing, or you know...

SH: Thrill.

JH: ...it's a big adrenaline rush doing that, so I'm very lucky that, you know, Shawn did kind of listen to me and say okay, you know, and get out of it.

RW: And now he's in rehab here.

JH: Well this has always been a place that we've loved and we never really thought that we would live here. But, and in fact we came for two months this summer intending to stay two months and piece after piece kind of fell in place. Our friend was wanting someone to take her house for the year and our son was in the middle of changing schools anyway and we just said well, why not us, why not we stay here and try it out. I think we're all ready for a change. I don't know how you can get more opposite than Silicon Valley. It's simple slow and, you know, a real uncomplicated lifestyle.

RW: Okay well lets hear about from the other couple involved in this scheme. So Kim who's this?

KH: This is Tracey my wife.

RW: Now are you the one that sells stock door to door?

TH: No.

RW: Oh that was somebody else. So what do you think about all of this business, company in Silicon Valley and all of that?

TH: Well I think it's definitely an American dream. You know, it's a, to me I'm coming from very different background. I came here 8 years ago from China and I went to graduate school here and I got first job in Silicon Valley in a company called Epic, which is a Meta competitor. So then I stayed there for three years and like the rest of engineers, you know, you changed jobs every two or three years. So I joined Meta that was in 1994 I think. I think I'm kind of lucky because that time Meta's already, you know, well shaped and bigger company than Epic, so I don't know, I really didn't know how hard Kim worked because you see him have a lot of fun there, so I really didn't go through that much, you know, the really small, the start up. But I heard, Kim told me in the hall that company started off in the garage, you know, work...

RW: Literally.

TH: Yeah, use peanut butter sandwich at the lunch. Kim would make, you know, peanut butter sandwich at lunch.

RW: Well that's great. How would you contrast this from China, Silicon Valley from China?

TH: Oh it's different. I don't think that anyone in China could imagine this, but now they probably do because they have a really lot of business going on, you know, off site investors going in and but, start ups. You see now everything is owned by the communist. Not like here, two people can start a company any time you know. If it didn't work, start again. But over there, you know, you do, now I heard you can start company over there but you know they need money first then you can.

RW: Well thank you very much.

KH: Thank you.