Lynn Conway's Retrospective
Lynn Conway [Draft of 2-09-05.]
Copyright © 1999-2005, Lynn Conway.
All Rights Reserved.
 
 
 
PART II: EARLY CAREER THROUGH TRANSITION
 
 
In June 1964, it was time for "Robert" to go to work by joining the prestigious IBM Research Laboratory at Yorktown Heights, N.Y. It was time of intense intellectual activity within the Lab, which was one of the most elite of computer research outfits at the time. I joined "Project Y" at Yorktown, working on the architectural concepts for a new type of extremely high-performance supercomputer. Then, in September of 1965, along with a dozen key members of the Project-Y architectural team, I was moved to California by IBM. There we began a crash project to develop the ACS-1 machine. ACS-1 was a super-secret high-performance supercomputer project initiated under the direct orders of T. J. Watson, Jr., the CEO of IBM.
 
At IBM-ACS, I participated in the innovation of what would become the "superscalar computer", which is of the cornerstones of modern computer hardware. I contributed the invention of generalized, multi-out-of-order issue "dynamic instruction scheduling" to the superscalar architecture. This invention went on to become a key component in most modern high-performance computer systems, including most PC's, as we'll see later in this retrospective. I also created the architectural level ACS-1 simulator, and guided simulation explorations to "tune" the new superscalar form of computer design.
 
For a few brief years, my family life was also sweet, especially the wonderful times with little Kelly, later joined by Tracy. These two little girls were a treasure and I loved them very much. They were the babies that I myself could never have. We spent lots of time together, especially in the out-of-doors, hiking, camping and wandering all over the beautiful California landscapes. But underneath it all Robert's marriage was a failure, and was doomed to early divorce. My gender condition had never gone away.
 
Finally, new medical knowledge appeared in 1966 that clarified the true nature of my gender condition. I also learned that full gender transition was now medically and surgically possible for cases such as mine. I was finally forced to confront reality, seek medical help, and complete my gender transition.
 
Gender phenomena underlie much what it means to be a human being. Our inner feelings and outer social identities are deeply imprinted with a vast web of gendered complexity, as socialization and culture interact with our inner natures. The deep personal meanings and complexity of one's gender are simply taken for granted by most people. These things are revealed only to people who face intense inner gender conflict. Those who then attempt to physically and socially change their outer gender confront all of that reality head-on, in a great rush of terrifying, bewildering, and yet often breathtakingly beautiful experiences.
 
Gender transition is also perhaps the last primal, deep taboo, and I was among the early pioneers here in the U.S. Given the way things were back then, my transition came at the price of being fired by IBM, being rejected by all my family and friends, being prevented from ever seeing my little girls again, and having to start a new career all over again in "stealth mode". This section tells the story of my four years with IBM, concluding with a sketch of my gender transition.
 
 
 
Play for more than you can afford to lose, and you will learn the game -
- Churchill
 
 
Fortune favors the bold -
-Virgil
 
 
 
 

 CONTENTS OF PART II:

 
3. Early Career at IBM Research, IBM-ACS
The ACS Project
Dynamic Instruction Scheduling
The Sweet Time - an Illusion of Normalcy
Evolving the ACS-1 Machine
Designing the ACS Design Process
Personal Failure and Angst
Climbing
Help Appears Just In Time
Confronting Reality
 
4. Transition
Project Cancellation!
Suddenly, I'm Fired By IBM
Completing my TS transition
Looking for a new job
 
 

 PREFACE

  PART I

 PART II

 PART III

 PART IV

  PART V

 PART VI
 
 
 
 
3. EARLY CAREER at IBM RESEARCH, IBM-ACS
 
 
In June of 1964, "Robert" joined IBM at the new T. J. Watson Research Center in Yorktown Heights, N.Y. This was one of the very top computer research laboratories in the world at the time. Athough IBM was a highly constraining, conformist "dress-code" type of company, the new research lab building was a spectacular place, and seemed to symbolize the significance of what might be going on there. I was thrilled to be working at such a center of computer research, and Sue and I were optimistic that this new job would also quickly ease our financial concerns.
 
 
Sue with little Kelly, outside the IBM Research Lab inYorktown Heights, NY
 
 
After several months of apprentice-like new member duties, I formally became a member of the computer architecture group under Herb Schorr. Herb was working on the so-called "Project Y", an experimental supercomputer design that was the precursor machine leading up to the ACS project.
 
I assigned was to work on methods for building a microarchitectural timing simulator for the machine. This simulator could be used to study exploratory architectural and register transfer level designs, in order to tune the designs and get leads on promising architectural possibilities. Soon after undertaking this new work, the project began to get lots of attention within IBM. Suddenly the company's Chief Executive Officer, T. J. Watson Jr., send down a personal directive to senior research managers at the Lab for a "go-for broke" supercomputer project, a super-secret Manhattan-project type of effort, to create the most powerful supercomputer possible at the time:
 
 "For more than two years, it has been apparent in the IBM Company that we were behind in the large scientific area. This is an area where, since the days of our Harvard machine, we have attempted to lead. Although four or five years ago there was some doubt as to whether or not we should continue to try to lead in this area because of the expense and other considerations, at some point between two and three years ago it became evident that the fallout from the building of such large-scale machines was so great as to justify their continuance at almost any cost. Therefore, for the past two years, under Vin Learson and Dick Watson, this subject has had the highest priority, at least at the upper areas of management of the corporation. In spite of this, our present 92 effort is merely a response to the challenges of the [CDC] 6800 as they have been announced. Almost never have we leaped ahead of them, but merely responded to their leads.... Now we have decided to take a new approach, set up a separate team, and go for broke on a very advanced machine in a very short time to come up with something so much better than the 6800 as to once more, in the eyes of the public, put IBM far away in the prestige league." -- T.J. Watson, Jr. (memo of May 17, 1965, emphasis added)
 
 
The ACS Project
 
The Project-Y team quickly became to focus of intense attention. We were the logical group of researchers to take on this new challenge, because the team was already working on many advanced methods relevant to such an advanced machine. I could hardly believe what was happening. I'd been with the company less than a year, and I was now on a high-level project initiated directly by IBM's CEO, Thomas J. Watson, Jr. This project, led by Jack Bertram, was the project that lager became known as IBM Advanced Computing Systems (ACS).
 
As on Project-Y, the architecture team was led by Herb Schorr. I was on that team along with Fran Allen, Jim Beatty, Ed Sussenguth, Don Rozenberg, Charlie Freiman, and the legendary John Cocke of IBM "Stretch" fame. We were the ones who would create the fundamental high-level system design of the new computer.
 
I was assigned the task of building the ACS-1 register transfer level (RTL) timing simulator, which was to eventually become the architectural-level "running-model" of the new machine. This simulator provided us with a spectacular advantage for creative explorations. With it we had direct access to a working simulation model of the evolving computer. All our architectural ideas could be evaluated quantitatively, and the results used to guide the evolution of the overall machine design. As the simulator builder, I was right in the thick of all the architectural work. This was an intense intellectual environment. We were full of excitement as team members produced new ideas for the rapidly evolving system architecture, and we checked them out via the simulator.
 
During this phase, IBM's legendary computer architect John Cocke frequented our discussions. John was quite a character. His role on the project seemed to be to gadfly about and stimulate thought and creative work in general. He was especially good at coming up with incredibly cool questions that tested existing designs and design paradigms.
 
For example, during the early work on the ACS-1, he asked us in a couple of meetings "how the heck can we get this machine to execute more that one instruction per machine cycle, on average?"* I obsessed about this question. I was so young and totally naive that I thought this was a design question, and that maybe I could solve it. I kept mulling the problem over, working on on the side. I kept trying to fully envision the controls of the initial ACS two-register scheme. I wanted to do this "visually" rather than by writing down a "control flowchart". I wanted to simulate in my mind's eye the movement of instructions through the machine under the constraints of control interlocks - trying over and over again to visualize the real constraints.
 
*Note: John didn't mention at the time that Gene Amdahl had written a paper a couple of years before claiming that the limit was "one" (I found this out later). John always wanted to test such "presumed limits", and this was an excellent method for generating really good research questions! I used it myself in my later research career.
 
In the summer of '65, the pace really picked up. We were now on a fast track. Herb Schorr, now manager of the ACS architecture group, spearheaded an intense architectural exploration to improve on the Project Y concepts. Herb had already done highly innovative work on the Project Y instruction set architecture (ISA), and he now worked out the extension of these ideas to the new machine's ISA (with Dick Arnold?). He also worked on the complex branch-handling mechanisms with Ed Sussenguther, and hit upon some fundamental innovations such as the branch history table. At the saame time, many others worked on redesigning and reconfiguring the pipelined arithmetic functional units and the bussing structures of the machine.
 
This was an extremely intense period of architectural innovation. We wanted to create a machine micro-architecture that could exploit multiple pipelined functional units way beyond John Cocke's earlier Stretch computer, and especially well beyond anything Control Data Corporation (CDC) was working on. This intensely creative team effort, led by Herb Schorr, was unlike anything I would experience again until I was at Xerox Palo Alto Research Center some 10 years later, while doing my VLSI work.
 
Then the news came down that we were moving to California! This was an exciting, appealing move to me, and it confirmed that I'd become a valued member of the architecture team.
 
I made a trip in early September to the Bay Area to do some apartment hunting. I quickly found a nice two-bedroom apartment just off Middlefield Road a bit south of Oregon Expressway. I then had some time to drive around and explore the area. Wow, there was such a different feeling in the air in California compared to back east. Something was brewing then, in the music and in the culture. I vividly remember driving over to the coast on route 84, and walking the wild beaches there.
 
Exploring the almost deserted beaches, and climbing around rock crags that separated one beach from another, I got really caught up in the wild beauty and strangeness of this place. Something stirred in my soul. I began to feel a wild sensuality again. I still treasure the song I first heard on the radio that day. It somehow captured the feeling of this new place for me: The In Crowd, by the Ramsey Lewis Trio.
 
I became incredibly excited about moving to California. There were many reasons, including a tug inside that this somehow had something to do with my gender situation. I couldn't admit it fully to myself exactly what that meant, but being near San Francisco somehow seemed to be nearer to freedom than I'd been back east. I made all the arrangements for the apartment, and Sue and Kelly moved out there soon after.
 
In the late summer of 1965, the architecture team had begun moving from Yorktown Heights to a temporary building on Keifer Road in Sunnyvale, CA. (We were eventually to move into a new building on Sand Hill Road in Menlo Park). I joined the team out there in mid-September. Keifer Road was a plain building, with simple partitioned spaces inside. It was a far cry from the elaborate offices at the Watson Labs in Yorktown Heights. However, if anything, the no-nonsense plain spaces there seemed to enhance the intense interactions among the team members, and at just the right time, just as the machine's microarchitecture was converging.
 
Dynamic Instruction Scheduling
 
Suddenly, in the middle of all this architectural work and the associated simulation studies, I got a clear view of the interlock constraints and hit on the "dynamic instruction scheduling" invention. The key insight quite literally came to me while I was taking a shower one night. I mulled it over and over in my mind and tried to visualize it - and then started diagramming it on paper. Gads - it looked like it really made sense. I quickly worked out the details of a special queue for holding and issuing multiple instructions, even out of order, per instruction cycle, subject to satisfying certain control interlocks. Using this very "regular structure", a very general scheme could be implemented in a hardware design employing simple matrices of lots of transistors to evaluate control interlocks in parallel. It would clearly be very fast, although costly in terms of the numbers of transistors.
 
The context and thinking process behind this invention was a curious confluence of fragments of switching theory, queuing theory and matrix algebra, things I'd just studied at Columbia. Ideas from all across this space commingled together in simple ways to make this thing work. The key seemed to have been to conceptualize what the overall problem really was, and then be open to anything I'd ever learned in order to patch together a clever solution to the problem.
 
I got very excited about this new scheme, and wanted to tell the ACS team about it. But I also had doubts. Is it really right? It seems so simple, so why hadn't anyone thought of something like this before?
 
One explanation was that the IBM folks usually did control-logic planning by using block diagrams of if/then sequences. I realized that this practice occasionally missed opportunities for parallelism. Approaching things that way meant that no one was "visualizing" the flow of instructions of data under the constraints of control. Instead they were merely doing formal calculations from prescribed inputs, and couldn't reach innovative solutions outside the envelope of the method of "calculation". Then too, maybe no one ever looked into this new zone even if they'd tried to do some more open visualizing, because it would "be obvious that it would require too many levels of logic".
 
However, the source-destination matrices enabled a great parallelism in the issuance-interlock checks, and thus got around the "logic-depth" problem. Although this parallelism required very large numbers of transistors, ACS had the charter to "go for broke" in the search for performance. Thus we had the right charter at the right time to exploit such novel methods. After all, this "large number of transistors" was only within one modest-sized component of the overall machine, namely the instruction-issuance interlocks. It didn't matter if that component used lots of transistors if it speeded up the entire machine by a factor of three or more!
 
So, I checked the design and check it and checked it. Gads, this thing really does seem to work. I then gathered up the nerve to present the DIS concepts and mechanism to members of the architecture team, asking for a team meeting to present them with some "new results" I'd just gotten.
 
There was quite a reaction to my presentation, almost stunned silence at first. Then everyone got really excited and there was immediate interest in a rapid evaluation of the scheme. I worked with several ACS logic design engineers on a rough but detailed implementation, and we confirmed that the general scheme would likely fit within the required number of logic levels (and would indeed be very expensive in transistors). Herb Schorr, Ed Sussenguth, Dick Arnold, Charlie Freiman, etc., all then enthusiastically supported a rapid, major revision of the ACS-1 microarchitecture so it included my DIS scheme.
 
As the ACS machine was restructured to include DIS, I worked with several other team members in early 1966 to write a detailed tutorial on the concept entitled "Dynamic Instruction Scheduling" (PDF). This tutorial was used to familiarize other architects, logic designers, packaging engineers, and compiler developers with the essence of the functionality of the new scheme, so that they could better reinterpret their own tasks in ways compatible with the new system architecture.
 
 

Figs. 5 and 6 from the original  DIS paper (PDF) of February 23, 1966:

   

 
DIS was thus rapidly integrated into the ACS architecture. The DIS name on the tutorial was the name of the "technique" or "functionality", but didn't stick as a name within the hardware. The new DIS hardware modules themselves ended up being called various nicknames by the team members and the logic designers, like "the contender stacks", "the instruction queues", etc. (what is now often called the "instruction window"). It all seemed like such a basic idea, and the modules slipped into the architecture so easily, that before long everyone on the project took DIS for granted.
 
I'd always been interested in the history of electrical engineering, electronics and now the emerging computing. I had a sense of how various key inventions had gotten embedded into those fields and then were built-upon and applied over and over again by others. I was particularly enamored of the key inventions made by the pioneers of radios, especially Edwin Armstrong - with his inventions of the regenerative radio receiver circuit and FM radio circuits.
 
I now became very excited, because I realized that I too had made an important invention. I just had a gut-feeling about it. I had an intuition that the DIS "circuitry" would someday be widely used in computers, that it would be an enabling invention, and that later generations of computer architects would have to know about it and would "take it for granted" in their designs.
 
Although the DIS revisions were soon taken for granted, the other members of the ACS team treated me very differently from that point on. I was no longer just a support person who was "programming the simulator". From now on I was accepted as a serious research contributor. It was a subtle yet profound distinction, and I could really sense it. It was a good feeling to be accepted in this way and get this respect from such a powerful group of researchers. This forever changed my life - I knew now that I could do serious stuff, and if I'd done it once I could do it again.
 
Unfortunately I never got the chance to talk directly about my innovations with John Cocke, and let him know that I'd cracked the instruction issue problem. He had stayed back at Yorktown Heights when we all came west. Although he came out later on to work on optimizing compilers for the machine, he kind-of walked around in his own little world and never seemed to learn or realize that I was the one who come up with the original DIS idea. He knew the general idea of the new DIS scheme, and later called it "a neat FIFO scheme". But to John I was still just "the kid who built the simulator".
 
The Sweet Time - an Illusion of Normalcy
 
During the first two years of Robert's and Sue's marriage they were totally preoccupied with all the changes in their lives: They settled in together, scraped to earn money, Sue gave birth to Kelly, and Robert started work at IBM. And it wasn't long before they moved on to California.
 
They had a lot to be thankful for. Kelly was a beautiful, loving little girl. They took great joy in caring for her and playing with her. They had a nice apartment in Croton-on-Hudson and Robert had an incredibly exciting new job with IBM. The move to California added even more excitement in their lives, and Robert's creative work intensified. They moved into a nice apartment in Palo Alto, California in September, 1965.
 
Sue, Kelly and Robert liked to spend lots of time in the outdoors, hiking in the Hudson Highlands and Adirondacks back east, and then hiking and camping all over California. Kelly loved to ride in a kiddy-backpack on hikes in Yosemite, along the California coast, and on the trails in the beautiful Palo Alto Foothills Park.
 
 

 

Robert and Kelly at their family campsite in Yosemite Valley
 
 
1966
 
 

 

Robert and Kelly on the trail to Upper Yosemite Falls
 
 
1966
 
 
It was a sweet time for all, and there was a sense of warm, loving companionship in the air. However, the marriage itself was an illusion. Sue and Robert both knew right from the beginning that something was wrong.  But so much was happening, and so much of it seemed wonderful, that they just didn't face the problems - at first.
 
Sue and Robert simply never had a passionate woman-man relationship. They hardly ever had intercourse, and even then only with difficulty and confusion.
 
Robert couldn't bring himself to talk about his gender woes, and sexual intimacy with Sue felt really strange to him. It seemed to him as if it they were two regular girls trying to have intercourse when there wasn't a natural attraction between them (except for Sue's longing for "Robert").  They both did enjoy softly cuddling together at night, and this seemed to create the illusion that there was a bond between them.
 
However, serious tensions began to build between them, because of Sue's frustration over the lack of sexual activity. He became distraught from the shame, humiliation and hopelessness of the situation and increasingly sought escapes through rock climbing adventures.
 
Robert and Sue had intercourse only about a half-dozen times during their entire four year marriage. However, during one of those rare sexual interludes Sue managed to get pregnant by Robert again (at least Robert wanted to believe it so), and in March 1966 she gave birth to another beautiful little girl named "Tracy". Everything appeared to be blissful to the outside, but both partners were becoming shattered by confusion inside.
 
 
[l-r]
 Kelly, Tracy and Sue
 
Yosemite Valley
 
 
1967
 
 
Evolving the ACS-1 Machine
 
Upon moving to California, we began a crash effort to quickly implement the new ACS-1 in the timing simulation system. This was quite and intense effort, and we just had to make it work and work quickly. The machine design was changing fast, and we had to quickly capture all those new ideas and structures in the simulator.
 
Herb Schorr, along with Ed Sussenguth, Dick Arnold, (and others?) did further pioneering work while redesigning the hardware branch handling mechanisms for the new version of the machine. As their design interactions proceeded, I worked on capturing their latest micro-architectural designs in the simulator, and on evolving and debugging the emerging design via simulation studies (see MPM Timing Simulation paper (PDF)). What was so cool about all this was that the simulator gradually became THE formal description of the latest version of the micro-architecture. This had great advantages, because any new design fragment had to "actually run" in the simulator, and we thus knew that the evolving design was really sound.
 
I also worked on iterations of logic design implementations of the instruction issuance interlocks, working with Merle Homan on fully confirming logic levels and machine cycle time constraints for the issuance numbers and depths of "window" we were settling on. I used the MPM simulator to measure performance on important benchmark kernels, tuning the machine's architectural parameters as we went along. We gradually zeroed in on a "3 out-of 8" contender stack sizing as providing the best overall performance tradeoff.
 
Later that year John Cocke came out to ACS, and we began running many simulations to support John Cocke's, Fran Allen's and Jim Beatty's exploration of software compilation methods for the new "superscalar" machine. The compiler team's work yielded another fundamental piece of the superscalar puzzle (along with my work on DIS and Herb Schorr's, Ed Sussenguth's and Dick Arnold's work on ISA, branch handling and exception handling), namely how to compile code from high-level languages so as to more efficiently feed it into this type of machine. John loved the new functional architecture of the machine (although he didn't talk with me and learn the details from me of how it really worked at the micro architectural level ), and he greatly enjoyed running simulation studies to explore new compilation schemes for the machine. John, Fran and Jim had come up with some important compiler innovations which they tested and evolved via these simulation studies. Thus they pioneered key superscalar software methods in parallel with the innovation and evolution of the ACS superscalar hardware architecture.
 
In a strange series of interactions, I worked with IBM patent attorneys in an effort to protect the DIS intellectual property. Herb Schorr and others encouraged these interactions. However, the DIS idea was such a high-level system concept in its day, and the logic design hardware implementations were so straightforward and consumed so many transistors, that the attorneys viewed the innovation as if it were a "software" idea, or a "mathematical idea", and didn't see a way to protect it! They myopically kept looking for a "logic design hardware invention". We went round and round on what it was all about, and nothing ever came of it while I was there. This was extremely disturbing to me, because I sensed that DIS was a fundamental new architectural innovation that would eventually have real impact (this turned out to be a good prediction).
 
I also got involved in a parallel very different activity. Ever since my studies at Columbia, I'd had a habit of playing "ethnographer". I enjoyed observing and noting the basics of what people were doing in little cultural niches such as hobbies, and small technical specialties. With the thought style of ethnologist and ethnomethodologist, I was always thinking "what did they think they were doing?". What fraction of their behavioral productions were imitative, and were efforts at "being accountable members" of their clan even when they didn't know what they were doing. I also wondered and speculated about where their practices, rituals and artifacts come from, and how did their rituals and artifacts channel their behaviors. These things enabled me to notice how niche cultural practices evolved and changed - basically "how they worked".
 
Thus it wasn't unusual for me to notice the clan practices of the various ACS engineering groups - things like what their work consisted of and how they went about it and how they "produced work". Suddenly, I noticed clear inconsistencies in logic symbol use and interpretation among the different logic designers, based on their differing backgrounds. What fascinated me was that the designers didn't even know when they were interpreting things differently and mis-communicating about things. However, it was clear that in certain subtle situations the logic symbols meant different things to different designers. I began to question what each designer's understanding was, and wrote this all down, designer by designer. I didn't tell them that I was questioning their understanding of what they were doing. Instead I tried to give the impression that "I wanted to learn everything I could about the IBM way of doing formal logic symbols". I then debugged this whole mess of data, and crafted an ACS tutorial that simplified and converged the various methods. It was entitled "ACS Logic Design Conventions: A Guide for the Novice", Nov. 29, 1967.
 
Designing the ACS Design Process
 
This experience led me to many more questions about the design process. How could we expect to get such a complex machine as the ACS-1 to work if fundamental human communication errors could occur at so many different levels of design, including architecture, logic design, physical design, physical implementation, and technology. What did these different teams know, and what didn't they know? I now realized that we didn't know how to put bounds on "what we knew" and "what we didn't know" about our design abstractions and their hierarchical composition. Buy if that were true, then how could we fully test heir productions?
 
All of these questions led me to begin work on the idea of "designing the ACS computer design process". I hoped to create a multi-level, coordinated simulation into which one could capture all the levels in the hierarchy of design abstractions. I crafted a specific proposal to ACS for how to design the design process (PDF), and it was very well received by the team. This proposal might very well have been implemented at ACS and used for the ACS machine, if the project hadn't later come unglued.
 
This experience in the conception of how to design a design process, and how to manage a complex hierarchy of design abstractions, had a tremendous influence on my later work in VLSI design methods. It enabled me to see that the levels of computer design were not "mathematically or scientifically" imposed but were "arbitrary cultural artifacts", themselves "designed" and potentially subject to innovative yet formal redesign.
 
Personal Failure and Angst
 
Even before Tracy's birth in early 1966, Sue and Robert's marriage was on the rocks. There weren't any fighting or emotional scenes, just a growing sense of total personal failure and angst. Sue talked endlessly with her girlfriends about her sexual frustrations. Robert picked up on this from time to time, noticing furtive
disapproving glances from those gals, and worried that Sue was seeing other men when he was away...
 
Sue and Robert both knew how serious the problem was, but avoided talking to each other about it. They both loved their little girls very much, and spent lots of time with them. At the same time, Sue and Robert drew further and further apart from each other. Robert began pushing himself ever more into climbing adventures, as a means of escape from the reality of this situation and block it all out.
 
Climbing
 
Moving to California had given me access to some of the best hiking and rock climbing in the world. I began spending many of my off-hours practicing at local bay area rock climbing practice sites, where I met other climbers and got hooked up with the local climbing scene. I also joined the Sierra Club, and took some of their climbing instruction, and then joined with a few others to form our own informal climbing group.
 
The Sierra Club had all sorts of hoops to jump through before they'd ever take anyone to anywhere other than local practice sites. Getting up our nerve, our little local band headed over to Yosemite Valley , guidebook in hand, to try out some of the easier rock climbs there. Although exciting at the time, we got up them fairly easily, and then just kept on working our way up to more difficult climbs. We all took turns at leading, i.e., at being the first one up a new pitch. The leader is the one who takes some risk while climbing, because they aren't protected by a rope from above as are those who follow later (we used pitions for protection back then, in the "old days" before chocks).
 
We had wonderful experiences over a couple of seasons, gradually doing the "old standards" during the summers in Yosemite Valley. Going to the Valley always got me pumped with adrenaline. The climbs there were on fine hard granite rock, and often involved a continually aggressive, physical approach. The protection was very good, so you could "push your limits of skill" without too much fear of getting hurt. But there was lots of exposure on the multi-pitch climbs there, so you could sometimes get into "mind problems". Sometimes we'd have great days and we'd all outdo ourselves. Other times mind problems would sweep through the team, and we'd chicken out, back off, and go work on familiar practice climbs instead.We learned many good lessons in teamwork, controlling fear, and showing grace under pressure.
 
During the winters, we shifted our climbing to Pinnacles National Monument , where on non-rainy days it was often delightfully warm and calm and sunny. I had very different feelings for Pinnacles compared to the Valley. The climbs there were shorter, but were on strangely brittle rock jagged, loose, metamorphosed volcanic rock. They were often done with dubious or outright poor protection. Travelling there I usually felt great, looking forward to the wintertime beauty and charm of the place. But once on the rock and leading at Pinnacles, it was easy to get the shakes, due to the poor protection, and the thought of getting ripped up by all that jagged rock if you fell. Therefore, at Pinnacles we tended to finesse things, taking it slow and cooly, and never "pushing things" like in the Valley. These were more good lessons in how to best take calculated risks.
 
These climbing adventures had a way of taking my mind off the looming disaster in my personal life, and my incredible fear and angst about it all. Having totally failed as a male, I feared that I would have to live all alone again, possibly for the rest of my life.
 
The camaraderie, the teamwork, the dance-like sensation of movement over the rock, the toning of one's body, the learning to take carefully calculated risks, the surges of adrenalin and sense of adventure, even just the smell of the all familiar rocks, all helped take my mind off things for awhile. As had often happened earlier in my life, I had a tendency to fall in love, quite surreptitiously and without ever showing it, with one of my "buddies".
 
This provided me with a fantasy escape on some of these climbing trips, when I could feel a close connection with a frequent companion, even though there was no way I could show him how I really felt (i.e., as a woman would feel towards him). Sometimes when these womanly feelings swept over me, they were in resonance with the sensual, emotional aspects of the climbing, and somehow fitted into that scene. However, it was with bittersweet sadness that all those feelings had to remain hidden and secret within me.
 
 
 Robert leading a couple of easy but rather photogenic climbs,
in Yosemite Valley (l) and at Pinnacles N.M. (r), '66-'67
 
 
However, I began to notice how the tourists, especially the girls, looked at us and admired us, and not only for our daring, but also for our handsome male physiques. Those girls weren't just looking at my climbing partners, but at me too. Robert was lithe and slender, nicely tanned, and had well-developed, well-toned muscles. Interestingly, part of the beauty of his body undoubtedly came from the lingering effects of Lynn's having taken female sex hormones during her early 20's.
 
Robert often had a stubble of facial hair when climbing over in Yosemite Valley. I guess he did that to look a bit more macho. Gads, this had become too good an act and had gone way to far! It wasn't just Sue who wanted Robert sexually, but other girls did too. I realized just how far his physical development as a guy had gone, and it began to shake me terribly. Now not only were my genitalia not right, but my whole body "wasn't right" either. I knew my chances at a life as a woman were slipping away, and I might now be forever locked inside a hardened male body that I hated, and with no escape.
 
 
Robert, resting after a day of climbing at Pinnacles in early '67:
 
A lot of guys would give anything to have a handsome, muscular male body like this,
 
but Lynn was near desperation at how the years and testosterone had ruined her chances to ever be a pretty woman.
 
However, Lynn was about to re-enter MtF transition anyways, and just hoped for the best.
 
 
Help Appears Just In Time
 
Then, in the fall of 1966, newspapers around the country had propagated the following item from a column in the New York Daily News:
 
"Making the rounds of Manhattan clubs these nights is a stunning girl who admits she was a male less than one year ago and that she underwent a sex change operation at, of all places, Johns Hopkins Hospital in Baltimore. Surprisingly, the hospital confirms the case, saying surgery followed psychotherapy. Such operations, although rare in this country, are neither illegal nor unethical, according to a Johns Hopkins spokesman. Officials at a number of major hospitals here agreed with Johns Hopkins on the legality and ethics of the operations but none could recall such an operation ever having been performed in New York."
 
I went on "high-alert" upon hearing this news, and followed the newspapers closely. Shortly after that, my life's trajectory was totally changed when the New York Times published an extensive front-page article on transsexualism, on November 21, 1966.
 
The Times article provided extensive information on the surgical and hormonal treatments then being done abroad, and on a new program at John's Hopkins University Medical Center, where several surgeries had recently been done. The article identified Harry Benjamin, M.D., as the leading authority on transsexualism, and author of a book on the subject entitled The Transsexual Phenomenon. Transsexualism thus presented seemed no longer just a back-street, porno type of subject, and the treatment possibilities were fully out in the open now . This article left me shaking and trembling, and my mind reeled as the full implications of the article settled in.
 
I immediately obtained and studied Dr. Benjamin's book. Dr. Benjamin was the pioneer of the whole new area of medical knowledge of transsexualism. His paradigm-shifting medical text described his experiences with many patients over the years. He was the first to recognize how gender identity and sexual orientation are two independent dimensions of each person's human nature, and he innovated methods for classifying the various types and intensities of "transgender" phenomena.
 
Dr. Benjamin also recommend how the rarer group of "intense transsexuals" could and really should be treated to enable them to live in the gender they sought. His book documented the results of the new, innovative surgical and hormonal treatments and put those treatments into a rational context as therapy for transsexualism. This book gave fresh hope to many transsexuals, and opened the door to the modern medical approaches that we now take for granted.
 

Confronting Reality
 
Finally the reality of it all came home to me. One evening, on a drive home from an evening with some other couples, I pulled over to the side of the road and began crying. The new knowledge I'd gained was almost too much for me to stand. I should have known all along that I'd taken the wrong path after the incident with the doctor at Boston University. Sue asked what was wrong, and I simply told her that I should have been a girl, and that I now needed to be a woman. She gasped, and seemed to instantly grasp that this was really serious.
 
I wasn't sure if she grasped the full depth of the situation. I figured that she might have thought I was saying I was "gay". But I knew I wasn't gay. Heck, if I'd wanted to be with men as a gay man, San Francisco was only 25 miles away and full of opportunities for that. No, that wasn't it. I needed to be a girl before I could find a man's love the way it was meant to be for me.
 
This disclosure led quickly into a very dark period. Thinking that time had passed me by and that it was now too late for me, I began to feel utterly hopeless, and then became suicidal. One evening, I collapsed on the living room floor, in the deepest hell of isolation and desperation. Not seeing any way out I loaded and cocked my .45 automatic pistol and put it to my head, sobbing. I was caught in the act by Sue, who had heard all the commotion.
 
Sue calmed me down a bit, and tried to find out more about what was wrong. For a while she seemed to think it was because of my "my performance problem and my failure as a guy with her". I tried to explain my situation to her, showing her Dr. Benjamin's book and telling her about the recent sex changes. I'd finally confronted reality myself: no matter how hard I'd tried or how successful I appeared to be as a guy, I couldn't escape from my innate gender identity. I wasn't a guy. I needed to become a girl, but I thought it was now too late for me and this thought was shattering.
 
Suddenly a lot of things became totally clear to Sue. In a long, emotional session, she talked me out of ever thinking about suicide, and pleaded with me to try to find help instead. Obviously I had to get out of this gender trap or it would kill me, because my life in its current form was simply over. And it would be best for her and the kids if I could work it out and find ways to continue helping support them.
 
Why hadn't this collapse happened sooner? Why hadn't I sought help earlier? I'd been seeing sketchy reports the past two years about incredible new methods for "sex change" surgery that had been invented by a french surgeon named Georges Burou. Using female hormones and these techniques, it was possible physically transition so as to legally take on a full female identity in California back then.
 
Not only that, but many patients were reporting that they were able to be fully responsive lovers with their men after their transformations. Dr. Burou's technique saved the patient's existing sensitive genital tissues and exploited those tissues during the construction of the new vagina and vulva. This technique usually eliminated the need for disfiguring skin grafts for lining the vagina, and greatly enhanced sexual feelings and orgasmic capability after the transformation. Dr. Burou was now operating a clinic in Casablanca, Morocco, and several famous "female impersonators" from the club Le Carrousel in Paris, including the beautiful performer Coccinelle, had gone to Dr. Burou and had wonderful results. I'd heard about those sex changes but didn't follow up to learn more about them.
 
Why hadn't I tried harder to find out more about all that? It was now obvious that the Boston University Medical School "expert" had been totally wrong when he'd said it "couldn't be done and I could never become a woman", but I'd bought into that and held on to it too long. I now believe that his opinions might have reflected the medical community's confused early reactions to Christine Jorgensen's initial surgery. Her surgery in 1952 had only removed her male organs; her vagina was constructed in a later surgery in1954, and some physicians may not have heard about that.
 
I had also been held back by the totality of the trap I'd fallen into, including my relationship with Sue and my responsibilities to our little girls. It all seemed overwhelming. Then there was the social outlaw image of the whole situation: Before Dr. Benjamin's book, the only information available was from sketchy underground newsletters and lewd articles and photos of "sex changes" in pornographic magazines. There was the fear, fear that it was too late, that I'd become too masculinized, and that I might not be able to make it work for me. And fear of how all my family, relatives, friends and colleagues would react, fear of the unknown, of how to go about it, fear that even if I could figure out how to do it, I might not have the courage to do it.
 
However, by now climbing had also helped me from rather timid beginnings to build lots of inner strength and confidence. I'd learned that if you wanted to do a climb badly enough, you had to step off the secure stances, off into the really big Yosemite Valley exposures, and gracefully, cheerfully make the needed moves in spite of sometimes terrifying fear and uncertainty. If so, you could make it to the top.
 
Doing this over and over again seemed to change me inside in some fundamental way. Those climbing experiences provided me, if not with courage, at least with powerful methods for calming myself and going into the line of fire no matter what. It readied me to calmly face what had to be done now.
 
In any event, by 1967 the hormonal and surgical techniques were developed well enough so that a complete gender transformation was possible for someone who was fairly young, reasonably healthy and not too masculine looking. I was 29 years old at the time. It was time to go do it. It was now or never.
 
 
4. TRANSITION
 
Thus began one of the most intense, dramatic periods in my whole life, my transition from being physically and socially male to being physically and socially female. Upon reading and re-reading Dr. Benjamin's book The Transsexual Phenomenon, everything came into sharp focus. It couldn't have been clearer what my problem was, and what I could and should do about it. All on my own, over many years, I had followed the classic pattern of the intense, early-onset MtF transsexual.
 
With Sue's support and full commitment to my transition, I'd sought out Dr. Benjamin's help and went to see him. He became very interested in my case, and decided to manage it personally even though he was semi-retired. I was probably one of the last patients he took under his primary care, as he was well into his 80's by then.
 
 

 Harry Benjamin, M.D.
in his "summer office" at 450 Sutter St., Suite 2232, San Francisco, CA
 
[this photo was taken by Lynn in 1973, five years after her gender transition]

 

[see also Dr. Benjamin's note in Lynn's copy of his book]
 
 
Back in 1967 we were still in the pioneering phase of this area of medicine. Dr. Benjamin's theories and recommendations were not yet well understood, much less accepted in the U.S. The surgical and social transformations were still very dangerous, calculated risks at the time.
 
Seeing a possible path out of my nightmare and into a real life, and willing to run those risks, I began the treatments to initiate a male to female physical transition. This involved intensive female hormone therapy, electrolysis to remove facial hair, counseling with Dr. Benjamin to prepare me for the challenges ahead, and making the many social adjustments in preparation for surgery and social transition. This was a very dangerous, risky path to take, a path full of physical pain and social humilations that had to be calmly endured. However, it was the only hope for me.
 
I made frequent trips to meet with Dr. Benjamin at his office in San Francisco when he was there during the summers. Dr. Benjamin started me back on estrogen,this time even larger injected doses of estradiol valerate than I'd taken before. He also helped me with my transition planning and preparations for surgery. During the winters I met with John Alden, M.D., at 490 Post Street in the city. A practical minded psychiatrist and counselor, he was highly recommended by Dr. Benjamin. Dr. Alden helped me face and adjust to the many difficulties of transition more calmly and confidently than I otherwise might have. My electrologist was Martha Foss, one of the very best in those days. Martha's office was at 133 Geary Street, Room 803.
 
All three of my caregiver's offices were within a few blocks of Union Square in downtown San Francisco, and that area has a special meaning to me to this day. I recall so well the many drives up the Peninsula, then down into town to the Downtown Center Parking Garage, and then the walks to Post and to Geary. On my weekly trips for electrolysis, I'd wait till I got to 133 Geary, and I'd then drink a mixture of tang and vodka to dull the pain before starting the treatment. I'd then usually sit out in Union Park for a while to sober up afterwards, before starting the drive back home. My caregivers were wonderful people whom I really befriended and who made a huge difference in my life.
 
As my treatments moved forward, both Sue and I gained confidence that my transition might succeed. Sue decided to hold off on divorce and stay with me until my transition was complete, seeing this as the only way to insure a good future for all of us. We both knew that we'd part afterwards, but hopefully I'd be able to continue my employment at IBM in some sort of position somewhere, and make enough money after transition to be able to provide a comfortable level of financial support for the girls.
 
 
Project Cancellation!
 
In 1968 terrible news came down from above: The ACS-1 machine was to be scrapped, canceled!!. The machine had gotten caught up in the success of IBM's System 360, and internal political pressure grew to make it S/360 compatible. However, that compatibility requirement meant throwing away all our "superscalar" inventions (as they later became known). This was a truly weird twist of fate. All our creative work and inventions, which had made such great improvements in normalized performance, had taken the ACS-1 machine's design beyond where it could ever be made S/360 compatible, but were now to be thrown away becuase of the success of the 360. It was as if our intellectual productions had actually doomed the project.
 
Worse yet, there wasn't any effort made to collect and organize what we had done. It all just seemed to go on the scrap heap, as if being on the losing side in a technopolitical battle meant that the knowledge itself wasn't worth anything. It didn't occur to me then that almost no one in IBM, and very few even within ACS itself, fully comprehended the significance of the innovations our ACS-1 architecture team had made.
 
This collapse of the project was an absolutely incredible disappointment. Not only was the machine canceled, but due to the secrecy, it was almost as if the project had never existed! I began to doubt that anyone would ever even know about what we'd accomplished! Some of the team members, those who identified as "IBMers", didn't seem too affected. True IBMers just didn't worry about much, as long as IBM took care of them. They just went on to the next project. Others, including me, were very deeply affected by the loss of our work.
 
What could I do now? Most of the architecture team immediately went back to IBM Research at Yorktown Heights, contented to move on to other research opportunties within IBM.
 
However, by the time ACS was falling apart in 1968 I was well into my carefully planned, well supervised medical transition. I'd thrown aside all feelings of timidity and personal modesty, and drove myself to do the very painful, tough things that had to be done. Sex change sure was not for wimps! Nor was it something you could do halfway. It was all, or nothing. Life or death.
 
I'd been on female sex hormone therapy, and undergoing extensive, painful, electrolysis treatments for the past year. Some things were already beginning to be noticeable: My facial hair was almost gone, my breasts were developing again, and quite rapidly this time. My body-fat was shifting and this was altering my physical contours. My skin was softening again very quickly, and I looked ever younger and "healthier".
 
Although I was now feeling much better, and these change were really wonderful to experience, I began to worry about how to tell IBM about what was happening to me. I dreaded losing my job, because the overall transition required considerable resources and was not covered by any medical insurance. By then I'd saved quite a bit of money, but it wasn't enough. I'd clung to a hope that things might somehow work out for me in IBM.
 
Suddenly, I'm Fired By IBM
 
Early in 1968 things got to the point where I needed to be open with IBM about what was happening. I talked firsts to the local ACS personnel people, hoping to figure out how we might resolve my future work situation. I prepared very thorough documentation about the medical nature of the problem, and how it was being treated. I gave projections of likely outcomes for various ways handling my employment situation. I sensed that this initially caused some local turmoil within personnel. My managers and several of my close colleagues on the ACS architecture team put in good words for me, and it looked for a while as if we'd worked out an arrangement for quietly continuing with the company as a woman. The plan was that after SRS and a period of recovery, I'd transfer under a new name to another IBM location in the Bay Area.
 
Then for some reason the personnel folks communicated my situation and our plans to the senior corporate management of IBM "just to make sure everything was OK". Suddenly it didn't matter what I'd accomplished at ACS, or how much intellectual promise I'd shown, or how much medical information I submitted, or that my colleagues tried to intervene and help. Just as the time neared for my sex reassignment surgery (SRS), I was informed that the deal was off. The top corporate executive committee had decided that they wanted me out of there. It is very likely that decision was made by T. J. Watson Jr. himself, and I was summarily fired by IBM.
 
I'll never forget the corporate rationale for firing me. The senior executives were convinced that "I would cause extreme emotional distress in fellow employees who saw me in the work environment if they ever learned that I'd been transformed". I've never been able to figure out what produced this viewpoint, but they were unwavering in holding onto it. They felt that they were "saving IBM employees from psychological damage. Maybe they just didn't believe that I could really transition, and thought I'd end up all messed up. Whatever their reason, this was bewildering to me, because none of the decision makers had ever met me.
 
Even more bewildering were accusations by IBM management about what a terrible person I was because of "what I was doing to my family". This was an incredible accusation for them to make, because I was trying desperately to transition in such a way as to HELP my family. After all, I didn't run off and abandon my girls, and use our money for my own selfish ends, as did many "real men" who ran off and deserted their families.
 
Over and over I kept hearing "how could you do this to your family!" Using some kind of twisted logic they considered me to be such a pariah that I deserved to be fired! I've often wondered if the executives who fired me ever realized that THEY were the ones who were hurting my family, by firing me and eliminating my income, the income we desperately needed to support our little girls.
 
There wasn't any way to reverse the firing. I was history. I hassled with them a bit and eventually they agreed that if I'd sign "resignation papers", they would give me a six months "consulting contract". Who knows, maybe someone in corporate asked the question, "what if, by some miracle, this pervert is successful?" More likely it was a very routine way of keeping fired employees quiet.
 
Sadly, my firing had many negative effects way beyond the immediate financial ones. The earlier construction of a plan for quiet transition within the company had helped me maintain at least some fragile ties with extended family, friends and colleagues, and their "acceptance" of what I was doing. To them IBM's acceptance up to now of what I was doing represented society's acceptance.
 
When IBM fired me, all my family, relatives, friends and many colleagues too simultaneously lost confidence in me. They became ashamed being seen with me, and very embarrassed about what I was doing. None of them would have anything to do with me after that. I even feared that most of them thought I'd been lying about IBM's initial support for my transition.
 
My whole world came crashing down, as I suddenly lost my job, my reputation, my extended family, and all my loved ones and friends. IBM had such an incredibly powerful reputation that being excommunicated by them meant excommunication from my profession, and from the conventional, middle-class, social world I'd been living in. It was analogous to a medieval person being excommunicated from the Catholic Church. Suddenly I was on the "outside", and a pariah at that.
 
Sue stayed with me for a while, right up through my surgery that November. We were then divorced in December.
 
Amazingly, when I left IBM in September 1968, no one ever asked for any of my design notes, papers and files. Who knows why. ACS was in disarray then, plus the company was so eager to get me out of there, that somehow they overlooked this detail. I'd done a lot of work at home, especially the last couple of months before leaving IBM, and had copies of most of my files there. I just kept everything, including my detailed ACS-1 hardware architecture notebook, and even the Fortran source code for the ACS-1 simulator. That work had meant so much to me, and I really believed that what we'd done there was very important. Thus I carefully archived everything and kept it all for years!
 
[ Note: the manner in which the ACS-1 project was canceled meant that the superscalar knowledge generated there emerged in fragments and propagated un-credited over the following years. The realization that members of the ACS-1 team were the primary innovators of superscalar knowledge occurred only very recently as historical reconstructions of the project got underway. See Note 2, below.] ,
 
 
Completing my TS Transition
 
Sex reassignment surgeries were not accepted by or done by the medical community in the U.S. back in 1968, except for a tiny handful of experimental cases at places like John's Hopkins.  Therefore, Dr. Benjamin referred me to two foreign surgeons, namely Georges Burou, M. D. (the pioneer of the modern SRS technique) in Casablanca, Morocco, and Jose Jesus. Barbosa, M.D. in Mexico.
 
Based on many factors, including the ease of returning for aftercare, I selected Dr. Barbosa to do my surgery. He was a prominent Mexican plastic surgeon who had an elite practice in plastic surgery at Calle 8A, No. 1987, in Tijuana, Baja California, Mexico.  Many U.S. patients went to him for cosmetic and reconstructive surgeries.
 
Dr. Barbosa was by then very experienced with Burou's penile inversion SRS technique for creating a sensitive vagina and female genitalia.  A number of trans girls from California had gone to him and gotten good results. I made a number of day trips to consult with him and undergo physical exams during August and September, 1968.
 
And then, in September and November 1968, I traveled to Mexico for the SRS surgery, which was done in two stages. The first stage was done on September 23rd, and on November 18th I went through the final irrevocable surgery, never once looking back.
 
Dr. Barbosa performed these SRS surgeries in a small, high quality hospital in  (Hospital del Prado, Calle Bouganvileas #50). The surgery took 4-1/2 hours, and I was in the hospital there for 2-1/2 weeks afterwards. Finally I'd completed that key part of the transformation from physically being a boy into being a girl.
 
 
Jose Jesus Barbosa, M.D.,
the plastic surgeon who performed Lynn's SRS
 
 
1968

 The small, elite hospital

in Mexico where Lynn

underwent SRS

 

1968

 

 
 
 
SRS was a very frightening experience given the surgical risks at the time. I had to go thru it all alone without any friends or visitors there with me. Fortunately all went well, the surgery was a success, the care in the clinic was wonderful, and I recovered without any complications. Three and a half weeks later I returned to the U.S., and calmly and hopefully took some time to do the things necessary to get ready to go find work.
 
By the next spring I began to feel incredibly better about myself. The degree of the physical transformation and the alteration of my perceptual setting was almost unbelievable. Within two months of the surgery, I'd healed well enough to begin experiencing my new physical sexuality. Freed of the confusions and angst of genital and physical mis-gendering, I was now able to become aroused and sexually warm a lot of the time.
 
It's so hard to explain all this to those who haven't experienced physical mis-gendering. Without reassignment, such a person isn't able to experience their full humanity, including their sexuality. It just doesn't work. Sex reassignment is NOT A CHOICE for such people. It is a life-enabling transformation that makes their physical sex consistent with their innate gender feelings. Without a consistent gender, one really doesn't have a life, much less liberty and a chance to pursue happiness.
 
The confirmation of this physical/neurological reality came soon for me. Without the horrible mental block caused by grossly inappropriate genitalia, I now became easily, openly and unashamably arousable. The remaining sections of the corpora cavernosa shafts inside my body easily became firm and hard, providing a powerful sense of "horny" sexual arousal and erection, and the surface nerves in the clitoral area above the entrance to my vagina were now extremely sensitive. At about two months after the surgery I began to have my first orgasms as a woman. I could easily get these orgasms by pressing down on and rubbing on a firm pillow between my legs, and these gradually built up to a point of intensity that was far more powerful than anything I'd been able to experience before. Sometimes I'd moan and scream as they peaked. This was a zone of ecstasy that I had not anticipated or even thought remotely possible. I was now experiencing a true second puberty, this time maturing as a girl, and all the wonders and surprises that entailed.
 
Now that the high-levers of testosterone were finally gone from my system, I could feel the full flush of the estrogen all by itself . Being on testosterone is like being on nicotine: One's nerves are always jangly and on edge, one is quick to temper, always ready to spar with others. But suddenly after SRS that "nicotine" was totally gone. Being only on estrogen, it seemed like I was sipping Gran Marnier instead. All the girly feelings were tremendously heightened, and there was a sweetness and an inner glow in my body that brought a continuous smile to my face. Have you ever noticed how all around the world young women smile and young men frown? It's not cultural. It's because of the dramatic difference in body feelings between being on lots of estrogen as opposed to being on lots of testosterone.
 
It was totally clear by now that ordinary everyday life as a woman would be ever so much fuller, sweeter, warmer and kinder to me than before, and least internally in my own mind and body. I now had a full, complete, warm and unhibited sexuality to bring to love relationships, and it has been that way ever since.
 
However, there were many uncertainties ahead. The internal mental, emotional and physical joy, warmth and peace came at a cost of many new external social, practical and financial difficulties. I was totally alone now, without any shoulders to lean on. The way people thought about such things in those days, I'd had to leave almost everyone I'd ever known completely behind, not just all my family members and relatives, but all my friends and acquaintances too. I had to go start a new life all over again, from scratch, like someone in what we now call a "witness protection program".
 
Let me tell you, this total social excommunication was terribly hard to bear. But there was no way around this back then. It was the common experience of others like me whom I'd came to know that even your closest friends couldn't find a way to stay with you, and adjust to watching the rapid progression of the physical and social transformation, especially given the incredible stigmatization back then.
 
If former friends saw you again afterwards and knew who you had been, the typical reaction was that it was all "make-believe", and that you were really still "so-and-so". You went through the gender "paradigm shift", but they never could quite catch on to the idea. Therefore, you just had to go off alone, get through the surgery and the transition alone, and then start a new life all over again with totally new people, like an alien orphan, alone in a new land.
 
When deciding on my new name, I made a clean break with the past there too. I'd used Lynn as my female name ever since my early M.I.T. days, but few people knew about that name now, so it was a safe choice. However, I needed a completely new last name. In late 1968, shorltly before my court appearance for my name change, I happened to read Helen MacInnes's novel The Salzburg Connection. I instantly liked the sound of "Lynn Conway", the name of the lead female character. Conway was a fine name for me because of my Irish heritage, so I took it for my own.
 
I knew that a totally new name would make it easier to "go stealth", and help me go right back into my original profession, right in the Bay Area. It would also help minimize the impact on people from my past life if I were ever "outed" later. Most of my family, relatives, friends and acquaintances had felt betrayed and embarrassed by what I had done and didn't want to be known as being connected with me. I sure didn't want to ever stir up that hornet's nest again, and a new last name would help avoid that problem.
 
In January of 1969, I was fortunate to be able to rent a room in the home of my electrologist, Martha Foss. She had a wonderful home at 1 El Sereno Court, on a quiet cul-de-sac high on the slopes of Mount Davidson in San Francisco. Her home had wonderful views looking out over the city, and a beautiful rock-garden down the back yard slopes. On moving in there, I made my social transition to living forever afterwards as a woman. This was the place and time for handling many critical practical matters, including getting my legal name change.
 
My attorney, Evander Cade Smith, a gay man who had handled a number of such cases, presented my case to the superior court in Oakland County. A particular judge at this court apparantly was familiar with transsexualism and sympathetic towards issuing name changes for those who'd undergone SRS. My appearance was very routine, their was no publicity, and it was over in a few minutes. By this time in California a legal name change with SRS documentation enabled a change in gender status on all other legal documentation. It also enabled me to get update my social security file, get my degrees updated, get a new drivers' license and get a new U.S Passport, all in my new name and gender.
 
Now that I was "properly documented", I needed to spend time out socially, and learn to be at ease in all kinds of new, practical social situations. It had been a long time since I'd tried to live as a girl. I was older now, and it was "24/7 and for keeps". This was awfully scary at first. On the other hand, I felt so isolated and terribly lonely being in the large house mostly by myself that I felt a strong need to get out and be among other people. I steeled myself for the possibility of harrassment, humiliation and insults, and began to get out pretty much every day. Fortunately, I had met a number of Dr. Benjamin's other MtF patients over the past two years, and was able to benefit from their practical knowledge of the San Francisco scene. During my transition I got to know eight other TS girls fairly well. Most were either entertainers (female impersonators in San Francisco clubs) or "working girls" in the sex trade in the City. Many of them lived in the Tenderloin district of San Francisco.
 
I picked up lots of "street smarts" from these girls, and often visited and socialized with them during my transition . Having some street-savvy TS girl friends to hang out with during transition helped me get out and rapidly build my confidence and extend my social horizons.
 
Tragically, two of my friends committed suicide during my transitional years. One was very beautiful young pre-op girl who slit her wrists after being gang raped by cops who had arrested her for prostitution. The other was an older preop TS woman who'd finally given up hope of ever making it through transition. She threw herself in front of a truck that was coming fast down one of S.F.'s steep hills.
 
But for me this was a time of rapid progress and hope. I quickly began to feel somewhat comfortable and relaxed about being out socially. My female secondary sex characteristics developed much faster and further than I'd expected after SRS (which had eliminated the sources of testosterone), and I was soon passing pretty well. I screwed up every once in a while, either in my voice or mannerisms, or by projecting fear and paranoia. I'd get "clocked" at those times, but it usually didn't cause much trouble other than weird, disparaging looks. Mostly though things were going better than I expected, at least going out and about among strangers. What worried me was how I would do in job interviews, and then in actual work situations where I'd have to see people day after day. I wondered whether I'd pass on an ongoing basis OK, or if people would gradually figure me out.
 
Looking for a New Job:
 
I began to run out of money, and ready or not started looking for a new jobin March of 1969, after only three months of recent social transitional experience. Fortunately, my attorney was able to hook me up with an experienced headhunter who had helped many gay people find good positions. My headhunter understood my situation, and the subtleties of how to handle "complicated placements".
 
My plan now was to "go stealth" once hired somewhere, in order to improve my chances for quietly "getting a life" and avoiding public stigmatization. I planned to disclose my situation behind the scenes to the HR department of a company once I received an offer from them. That way I hoped to sign up with a company that was OK with my situaition. Then even if I were to be "outed" anonymously to personnel by somebody later on, they would already know about my situation and know that I wasn't causing anyone problems.
 
[ Note: This turned out later to have been a good decision. To my knowledge, I never was "anonymously outed" to personnel in any of my jobs. However, by avoiding lies or misrepresentations on any prior job applications, I was in a solid position to get a Top Secret security clearance for high-level work in the Department of Defense later in my career, in 1983.]
 
I managed to see Sue, Kelly and Tracy a few times that winter, about once a month. Sue insisted that I had to try to look like "Robert" during these visits. It was increasing difficult to do this, and it this was incredibly stressful for me. I missed my little girls terribly, but dreaded the embarrassment of trying to look like a guy when visiting them. We didn't have a clue about how to handle this situation, and there was no guidance back then about what to do or expect. Some people thought it would "psychologically damage" children to even hear about gender changes, much less see one happen. Others thought that kids could adjust quite easily to gender changes if they were treated matter-of-factly, even more easily than adults (we now know that this latter is the correct prediction). We became stressed out trying to figure out what to do, and also very worried as our money began to run out.
 
I interviewed at a bunch of places that spring, and in most of them I got right up to the point of getting offers. Then I'd have to deal with detailed medical questionnaires that would "out me". I never wanted to misrepresent myself, so in each case I told the truth on all forms and questionnaires, and one right after another, job offers went up the corporate personnel ladders and then were rescinded. Again these negative decisions were made by people who'd never met me, and who didn't want to meet me, having some horrible stereotype in their minds instead.
 
One particularly disappointing example was the RCA computer research lab in Palo Alto, where there was a great job match and they really wanted to hire me. When I told them about my past and showed them all my medical documentation, they assured me it would be no problem. They then checked with the corporate level at RCA, and back came a request that I see Julian Pichel, M.D., a very prominent Palo Alto psychiatrist, for an evaluation. I went to the interview and thought it had gone very well. However, I was stunned when the company later called and said they couldn't hire me.
 
A few weeks later, Jack Leer, the personnel manager at the RCA Palo Alto labs, sent me a surreptitious copy of the Dr. Pichel's report: It was a wonderful report, full of extremely favorable comments. The report solidly indicated that there was nothing in my situation that should preclude employment in my field. The company had chosen to completely ignore the report, and declined me while giving the impression that the interview had been unfavorable. Fortunately, Dr. Pichel's report proved useful over and over again in later job searches, so the RCA episode turned out to be a blessing in disguise, for which I am forever grateful to Jack Leer, wherever he is.
 
Such episodes began to erode away my confidence, and by May 1969 the situation was becoming very desperate. I'd run out of money, and was having to borrow from friends in order to put food on the table. Not having money to do anything, I began to spend my days alone at home, waiting for the phone to ring and hoping that one of the many jobs I'd interviewed for would finally come through.
 
I even wrote one last appeal to Dr. Gene Amdahl who was then Director of IBM-ACS, pleading with him to help me find some kind of position, somewhere anywhere. He replied that as an IBM manager he could not help me, since the earlier IBM decision by the senior executives represented IBM's final position on the matter.
 
At the same time, Sue had had to go on public assistance for three months. She was working now, but there wasn't enough money and I'd just missed two months of child-support payments. In the process, the county social workers found out what had happened to me, and my fate was sealed regarding ever seeing my little girls again. Suddenly, everyone began to hound Sue. Her Catholic family and relatives, all her friends, and those meddling social workers all considered me to be such a terrible sexual pervert that it would "irrevocably damage my children" if they ever found out what I'd done. I was finally confronted with being arrested by the police if I ever tried to see my little girls again.
 
The problem of course was that everyone saw a "boy with a terrible mental problem" rather than a "girl with a terrible physical problem". That simple difference in perspective makes all the difference in the world in whether someone can grasp the reasons for transsexual gender transition. However, it is a perspective that most can only achieve if they first meet someone AFTER they've fully transitioned, and not before. It is then difficult to visualize the "boy", because only the "girl" is now visible.
 
The ensuing enforced separation from my girls was one of the hardest things I've ever had to bear in my life. I loved them as if they were my own babies, so it was absolutely crushing to never be able to see them again. For years and years into the future I would often cry myself to sleep at night thinking about them, but there was nothing I could do about it. They were gone, almost as if they'd been kidnapped from me.
 
I've often thought that if I hadn't been fired by IBM and that terrible extra stigma hadn't come down on me. The social workers might never have become involved. We might have had time to work out a more humane way to handle the situation. As it was, not only I but also the girls suffered greatly for many years [see Note 1.]
 
Then, just when things seemed darkest, I finally landed a job. In May 1969 I began work as a contract programmer at Computer Applications, Inc. (CAI). My hiring was a local responsibility by people who had met me, rather than a remote corporate decision. That made it a lot easier to get hired, even though the personnel people knew the story.
 
I packed my few belonging into my VW, left Martha Foss's home in San Francisco and drove on down the Peninsula to Sunnyvale. I'd was running on borrowed money by now. Having to settle for the lowest cost place possible, I moved into a cheap one-room studio apartment near the U.S. 101 freeway. However, my new job at CAI meant I now had a chance to make it back into society. I hate to think of what might have happened if that job hadn't come along when it did. This had been a close brush with total personal disaster , but I'd made it after all [See Note 3.]
 
 
 
Note 1. Even though under threat of arrest if I tried to contact or see my little girls, I had to pay child support during the years of their minority. I was eager to help, and willingly paid that support, but it seemed incredibly unfair that I could never see them. There wasn't any way to contest this treatment. Legal action was likely to fail, and such efforts would have outed me in Bay Area papers and cost me my new career too.
 
Within a few years Sue remarried, although she never told me about that. She never told Kelly and Tracy what happened to me either, and the girls grew up with a profound sense of loss and separation that was clouded in mystery. When Tracy was 15 years old, she finally confronted Sue and demanded to know "where her dad was". Sue's response was to draft a written letter of explanation and give it to Tracy, who then showed to Kelly. She never talked directly to the girls about it, then or later. Such was the shame and embarrassment regarding my transsexualism that Sue had internalized from all those around her. On the other hand, by quickly putting our relationship behind her and finding someone else to marry, Sue at least had been able to get on with her own life.
 
When Kelly and Tracy each reached age 18, they were finally freed from the "protective" constraints against contacting me and seeing me, and we began to write each other. In 1985 they both visited me at my home in Ann Arbor, Michigan, and we have since become fully and warmly reunited. Since they didn't remember Robert at all, they don't mis-gender me now. I'm probably closer now to Robert's children (as their Aunt Lynn) than most parents are with their adult children. By the way, Kelly and Tracy are both happily married; Kelly has two boys, and Tracy has three boys and a little girl.
 
 
Note 2. Although the original ACS project was canceled and the ACS-1 machine never saw the light of day, my key invention was so noticeable and fundamental that it survived and propagated by a number of means. Dynamic instruction scheduling, i.e., multiple issuance of instructions, out-of-order, per machine cycle, subject to appropriate interlocks, has since become a classic hardware method for enabling instruction-level parallelism (ILP). DIS is now at the very core of the architectures of many important VLSI superscalar processors, such as those by Intel, Sun, MIPS, HP, Compaq, etc., and greatly enhances the performance of those processors.
 
In 1999, I learned a lot more about what happened back in '68 from Prof. Mark Smotherman's ACS historical reconstruction. A group led by Gene Amdahl had been the winners in the political struggle to make an ACS machine that was 360-compatible. Ed Sussenguth became manager of the architecture group for the new machine. Efforts were made to integrate the primary superscalar innovations into the ACS-360 design, but how successful these were is not clear. The 360 instruction set precluded effective superscalar-type branch-handling, so even if DIS were included in the design it is unlikely that the ACS-1 level of performance enhancements would have resulted. The ACS 360 project itself was very short-lived; it collapsed only one year later when Gene Amdahl left IBM to start his own company.
 
ACS-360 team members later applied for patents on various IBM-ACS innovations. On December 30, 1970, Leo Hasbrouck, Bill Madden, Robert Rew, Ed Sussenguth and John Wierzbicki filed for a patent that described the ability to issue one or more floating-point instructions, out-of-order, from a buffer to one or more functional units. This filing obliquely revealed the dynamic instruction scheduling concepts, and attempted to cover them. It was issued as U.S. Patent 3,718,912 on Feb. 27, 1973. Although it included several diagrams of the form I'd used to teach the DIS concepts within ACS, it made no reference at all to my 1965 ACS-1 invention of DIS or to the 1966 DIS tutorial.]
 
Reflecting back, it's now clear that even though the ACS project was thought of as very secret within IBM, a number of high-level outside computer experts became familiar with the machine's novel architecture, and those experts began to diffuse the key knowledge out from ACS in the mid-late 60's.
 
For example, in part of the fascinating Lawrence Livermore Lab career retrospective interview of computer scientist Norm Hardy (23 May 1994), Norm reflects on his interactions with people at IBM-ACS in the 60's, and talks about the out-of-order multiple-issuance of instructions (DIS) in that machine. Here's a clip from that interview: " It was superscalar, and the instructions were rather simple - - it was the first design that I had heard of with the idea of issuing more than one instruction per clock cycle - - I can recall thinking that that was barely credible at the beginning. It turns out, in retrospect, to be the right way to build a machine. But I can remember being quite in awe of the very idea--in fact, I thought that executing one instruction per clock cycle was quite remarkable - - ."
 
In addition, that December 30, 1970, patent filing by ACS-360 staff members obliquely revealed the DIS concepts, and attempted to cover them, in the patent that issued as U.S. Patent 3,718,912 on Feb. 27, 1973.
 
Then, as we'll see later in this retrospective, I also deliberately propagated the superscalar knowledge out into the emerging VLSI system architecture community in the late 70's, again without attribution. As time progressed and the knowledge diffused further in the late 70's and early 80's, many heard of it without knowing the source.
 
Since no one ever claimed to be the innovator of DIS there was a void surrounding this important computer architecture invention, and others began to make such claims. For example, Prof. H. C. Torng of Cornell University claims to have invented DIS in 1982. He was awarded a patent which was pumped up by Cornell's pubic relations team so that it really seemed to be the original general DIS invention. Cornell eventually sued Intel for patent infringement, and Intel caved in, not knowing of the prior ACS art. In 1997, Cornell and Torng got an $8 million dollar settlement, and Torng was named as "the first Intel Academic Research Fellow" (see Cornell Press release regarding this settlement).
 
In addition, Prof. Y. N. Patt (now of the University of Texas), after designing a machine containing DIS in the early-mid 80's (the HPS machine), gradually allowed the impression to spread that he had first invented general, out-of-order, multiple issuance of instructions in that machine design (Patt contrasts that work with the earlier Tomasulo method, which was only single out-of-order issue). Over time that misleading impression has stuck, and Patt has won many top awards for that work. He now openly claimed to have invented DIS and the superscalar architecture. Coincidentally, Prof. Patt received his Ph.D at Stanford in 1966, right at the time the DIS/Superscalar knowledge was beginning to diffuse out of very nearby ACS (recall Norm Hardy's comments above).
 
However, the recent efforts by Dr. Mark Smotherman at Clemson University have finally led to the open compilation of accurate historical details about the ACS project, and to the confirmation that the ACS-1 was the first true "superscalar" computer architecture. For further information, see Dr. Smotherman's extensive ACS historical retrospective at: http://www.cs.clemson.edu/~mark/acs.html Details regarding my dynamic instruction scheduling invention appear in the ACS "Technical Description" subsite at: http://www.cs.clemson.edu/~mark/acs_technical.html.
 
Many further details are contained in my personal ACS archives, including my internal IBM-ACS papers, and even the source-code for the ACS-1 RTL machine simulator. I have sent those archives to Dr. Smotherman, for his use in reconstructing the details of the ACS-1 machine and the overall ACS project. The front-matter for my ACS archives lists the contents and describes some of the context of each item in the archive.
 
Up till now, I've very seldom talked with anyone about my life and work before 1969, including my work at ACS. I was busy getting on with my life, and I didn't want to call attention to things that would put me under constant scrutiny, especially back when people were far, far less understanding than today. However, as DIS became so widely used in VLSI microprocessors during the early 90's, and as others began to claim that invention, I suffered a lot of angst from having to remain silent. Ironically, it was the historical resurrection of the ACS project in the late 90's that finally created the need to be open about my life before '69, since my archives turned out to be the main source of information that confirmed the details of the ACS-1 machine's micro-architecture and hardware design.
 
 
Note 3. Never in my wildest dreams would I have thought, when starting over and trying to get that first new job in 1969, that by 1980 I would have established an international reputation for VLSI research work at Xerox PARC, or that this VLSI work would then help provide a context in which the ACS architectural knowledge would later flourish, or that I'd live to see much of that knowledge embedded in huge numbers of VLSI processors in the PC's of the 90's, and that my archive would then finally be used to begin a historical reconstruction of the ACS-1 machine in '99 to find out where the knowledge came from. But we're getting ahead of ourselves here. That's "the rest of the story", to be unfolded.
 
 
As a "teaser", here are some photos from 1999:
 
 
 
 
IBM-ACS Mini-Reunion,
IBM Research, Yorktown Heights, NY
 
[ left to right ]
John Cocke, Fran Allen,
Herb Schorr and Lynn Conway
 
July 24, 1999
 
Photo by
Mark Smotherman
 
 
 
 In memoriam:
John Cocke, 1925-2002
 
 
 Lynn Conway and Maurice V.Wilkes
 
HotChips 11 Conference,
Stanford University
 
August 17, 1999
 
 
Photo by
Forest Baskett
 
 
 
 
 

 PREFACE

  PART I

 PART II

 PART III

 PART IV

  PART V

 PART VI