Dr. Frank Soltis on the Birth of AS/400
Paul Tuohy talks to the father of the AS/400, Dr. Frank Soltis in part 2 of a three-part conversation. In this installment, Dr. Frank talks about the early history of the AS/400 and how it came to be.
Paul: Hi everyone and welcome to iTalk With Tuohy. This is part two of the lengthy interview I did with Dr. Frank Soltis. And in this part of the interview, I talk to Frank a little bit about the history of the AS/400 and about what was probably one of the greatest skunks works projects that ever happened in IT. I hope you enjoy.
Paul: Yeah. Well I have got a really tough question for you, Frank, because if you look back over your phenomenal and illustrious career, is it possible for you to pick out one or two highlights that you look back sort of going like they were the wow moments for me?
Frank: Well one or two. Let me think about that. Yes, I guess one of the things that really amazed me—well let me back up in time a little bit. In my early career, I was what I would today call a lab rat. I mean I lived in the laboratory. To me, it was great. I would get up every morning, go into the lab, work during the day and, you know, meet with a lot of people, design things, and have a lot of fun. You know the only thing I ever did to get out the lab is well I traveled to other IBM locations, IBM headquarters, or a few things like that. Once in a great while I would go to a conference and it would be one of these really pointy headed technical conferences, you know the IEEE computer conference or an ACM conference or whatever and I would actually present a paper of some kind, whatever it was but during that early period in my career and I want to say almost for the first 20 years, I never saw a customer. The number of times when I actually spoke to a group that had customers in it I could probably count on one hand. So up until we announced the System/38, I had done very little of that and I was greatly encouraged to attend COMMON in 1988 at which we were going to—you know make the announcement. So I will not say reluctantly. I was not all that excited about it but I said okay, I will go to COMMON because people were telling me you really need to see, you know, a user group conference. Right? I had never done that, so I walked into COMMON, and I was blown away. I do not know if you remember that particular conference. I mean it was huge, you know the number of people that we had. It was in Toronto and I mean it was monstrous. This may sound kind of stupid but it sort of hit me, you mean all these people really use this system? (laughter) I mean it really did. You know it was that wow moment that—I mean you sort of intellectually understand somebody is using these because somebody is buying them but you just never see it. That level of enthusiasm, the number, and the excited people about, you know, it was the AS/400 announcement. It blew me away. It really did and at that point, I mean I have done an awful lot of customer conferences since then but that was really the moment that it hit me that hey, this is—
Paul: Yeah, it’s that moment where suddenly a number has a face.
Frank: Yes. Exactly. That is the best way to describe it, absolutely because I always saw the sales numbers, you know these are going out. We are building bigger ones, etc., etc. but you just don’t like you say, you don’t put a face to it.
Paul: So anything else, Frank, that was a highlight?
Frank: Well I think, you know, there have been a lot of highlights. As we started, you know, doing some of the announcements, we start doing things out, just meeting with customers over the years. You know you kind of sit back and say wow, this is really pretty impressive stuff right? People sometimes have asked me, they say what are you most proud of? I say the answer is very simple. I am proud of the customer satisfaction. You know if I had to rank one thing of the systems, the iSeries up to IBM i, I would say it is customer satisfaction. Over the years, other counterparts if you will at IBM, you know, responsible for other systems in IBM have always said to me over and over again we would trade places with you instantly as far as your customer satisfaction. You know that to me has been, you know, one of the real positives. Yeah, there are things wrong. People complain about this, that, or whatever but by and large no matter where I go in the world people are satisfied. They are happy. They say it is a really great system and oh, by the way if you would only do thus and so, you know it would make it perfect right. So that from a highlight standpoint there has been just untold numbers of those.
Paul: Yeah. So to go the opposite way then, Frank. If there were sort of highlights, were there down moments? Were there times when you were going oh, to hell with this?
Frank: Well yes.
Paul: Okay. I should preface this now. Is it possible for you to share these moments with us?
Frank: Oh, oh, okay. Probably one of the biggest down time was in 1983. Okay we announced the System/38 in 1978. It was delayed in shipping because we had not finished making it work and so we did not ship it until 1980. Remember this whole thing in IBM about this five-year life of systems? It was 1983. I remember it well. I was summoned to corporate headquarters for IBM and, you know, meeting with our top executives including our CEO and it was one of these where this has been a wonderful—System/38 has been a wonderful product. It has done quite well. It is, you know, going to be five years old, you know it was kind of in that timeframe. Our executives said we sort of anticipate that sales of this are going to start falling off because, you know, so many years as I knew you have to come out with a new model of the system. They wanted to share with me their vision of the future and the problem that they were faced with. We were all in the same—we had just put all of our midrange systems into a single division. It was a new division and we had five midrange systems within IBM at the time. Two of them came out of Rochester. There was, you know, the System/38 and the System/34 soon to be the 36 and there was the low end of the 370. There was the Series I. Series I never started as a midrange, you know, business system but they migrated into it. Then there was another one called the 8800 system. They were all in the same division now. We had all been put in the same division and they said look, there is absolutely no way that we can continue to build five systems so we are going to merge all five into a new system and that is going to take over so when the System/38 sales roll off, this new system will take over. What is this new system? What are they planning? Well it is a 370. Now during the 1980s, the IBM Corporation had a major problem that the executives believed everybody wanted a mainframe and they put all the emphasis on mainframe. I mean the downturn in the early ’90s when, you know, IBM literally almost went out of business, had to go out and hire Lou Gerstner and so forth was all because of the total focus on the mainframe during the ’80s. In 1983, that was kind of the beginning of it and so the idea that they had that was going to be wonderful is they were going to have Rochester since Rochester will not have anything to work because they are getting rid of the 36 and 38. The 36 had not quite been—they were going to announce the System/36 at that point so IBM Rochester was going to get the opportunity to write a new operating system for the 370. Wow. Just what they needed. By the way, they had six operating systems on the 370 at that time. I don’t think I can name all six but there were six of them so they were going to get to build #7 right. They needed somebody to head the architecture group for that new operating system and I was offered the job. Well I was polite because of the audience that I had and said well let me think about it but my response was, are you kidding? You know so I came back to Rochester and said to my wife well it looks like—you know what do you want to do next kind of thing. I was really, really down on it because I said I do not want any part of building on the 3—I mean I was dedicated to this whole System/38 job. So I get a call from the lab director, a fellow by the name of Tony Mondello, which maybe not too many people remember Tony but he was a great guy. He calls me up, says come on in, and talk to me. I went to see him and he said look, I am putting together a new organization. Rochester needs an advanced technology group to work on really advanced things and he says I want to put this organization reporting directly to me right and I would like you to come to work in that organization. I would like you to manage one of the departments in fact. He says I want you to manage the system department and pick and choose some of the people that you want in the laboratory. We are going to do new advanced things on the System/38 and the System/36. So to kind of hide the advanced technology group, you remember how Rochester is laid out—we got the blue buildings and then there are the white buildings across the way. On the main floor of the white buildings was the customer center there, you know, the briefing center. Well on the lower floor, which was kind of the basement of that, was the advanced technology group. We took over that lower floor. So I said yes, let’s go do it. We had some of the most fun times during that period of time, during the ’80s working on some crazy things for both the System/38 and the System/36. We had most of the top people in there. I mean we had guys like Dick Bains who was the language guy who was working in that area. We had Dick Mustain—I do not know if you ever met—do you know Mustain at all?
Paul: No. No.
Frank: He was more or less the chief architect for the 36. On and on. We had the top 36 people; the top 38 people in that group so we are sitting around saying, you know, they are trying to merge these things together. Meanwhile there was some operating system group running in Rochester. They brought some new management in to head it up and all kind of crazy things and they were going to try to build this new operating system for basically the 370. So they were off doing their thing. We didn’t care what they were doing—
Paul: That was—sorry to interrupt. That was Fort Knox. Was that the project?
Frank: Yep, that’s what it was. Fort Knox was the whole division. You know other groups were working on other things but yes, that was Fort Knox. So we are sitting around, kicking around some ideas and somebody said why can’t we run System/36 applications on a 38? Okay. Well there is no reason we can’t right? We have got to make a few changes to it and things like that, so we could. So we started this funny projects, skunk works project called Silver Lake and people kind of know what happened there. By the way, IBM did announce that one 370 that the world wanted. The philosophy was that the only reason a customer bought a System/38, 36, or 34, whatever is because they could not afford a mainframe; they could not afford a 370 and if they could get the price of the 370 down low, these people would flock to it. Well that low-priced 370 came out in ’87. I think it was ’87; I have to stop and think about the year was. I believe that is when they actually announced it. It was called the 9370. Remember that one?
Paul: Yep. I do indeed.
Frank: Uh-huh. It was a good piece of hardware. I mean it really was. It was a well priced everything as far as a piece of hardware. The problem is that you needed all the expertise of running a mainframe and it was just totally unsuccessful. In fact, it was pretty much telegraphed in the mid ’80s that it was going to be unsuccessful so all of a sudden our efforts to say well can we merge a 36 and 38 all of a sudden got a spotlight. The head of the division who made these decisions was replaced. Steve Swartz came in as the new head of the division and, you know, immediately put his guy if you will as the lab director in Rochester. That was—you know Tom Furey came in and full speed ahead on what became system—the AS/400.
Paul: Yeah. You are bringing back memories for me. Frank. I remember that back in the ’80s because I was just like working for a customer at the time and I didn’t have the connections in later years at IBM at the time but I remember at the time depending on who you talked to in IBM, the future was going to be either the 370 line or the midrange line, you know what was later to become the 38 stroke AS400 line. It really was. There were these two camps of people in IBM and you would get a different story depending on who you talked to.
Frank: In fact absolutely. In fact, it was really kind of strange because one of the code names, the one that, you know, most people knew of what became the AS400 was called the Pacific. A lot of people knew that, right? The other group, the 370 group, came with their own code name called Atlantic, which very few people know about. It was not all that successful so yes, when all of a sudden it was the two groups in competition with one another. That is exactly what was happening. When the 9370 was announced and like I say, I know some of the engineers who work on the 9370 and they created a good piece of hardware. There was no question; in fact, a lot of that hardware was used in the AS/400. That was the big white boxes, you know the frames and a lot of the internals, but it just was not going to fly because, you know, from a user standpoint you really had to have, you know, the staffing and everything like that to run a major mainframe operation. So that to me, the whole thing, that was really the lowest point. It truly was.
Paul: Yeah, I just love the way though, Frank, that you just turned the lowest point into the birth of the AS/400. (laughter).
Frank: Well it was not—it was not me alone by any means. I mean I look back at—you know I look back at Tony Mondello, the lab director I mentioned, who had the foresight to say I am going to put a group in place to continue to focus on the 36 and the 38 because I don’t buy into this stuff either and Mondello came out of the mainframe group. I mean he came out of Poughkeepsie. His background was in MVS, he came into Rochester, you know, appointed as the lab director, and what is this small thing called the System/38? They began to show it to him and he went wow. This is what the mainframe should be so I mean he was just in there so, you know, you had to have the people with that kind of foresight to say we are going to put this group in place and here is where we are going to go. Right? He kept us totally isolated. I have no idea how it was funded or anything like that. You know that was not any of my business right and every once in awhile some top techie guy from Rochester would walk in the door and said well I got myself kicked out of the other project. Do you have any jobs here? (laughter) Literally, that is what it was because we were in a separate building. We were the outcasts. We really were but that is how it all happened.
Paul: Well that is it for part two of the interview. Make sure you tune in to the next iTalk when I will be talking to Dr. Frank about one of the other great passions in his life and a strange thing about how systems get named. Talk to you all then. Bye.
Like what you just read? To receive technical tips and articles directly in your inbox twice per month, sign up for the EXTRA e-newsletter here.