Part 1

[Section head­ings cor­re­spond to items in the video playlist, as this pre­sen­ta­tion was bro­ken into six pieces.]

Introducer: It’s an hon­or and a priv­i­lege to intro­duce today’s ple­nary speak­er, Frances E. Allen. Like many women of her gen­er­a­tion, Fran Allen start­ed out her career with a degree in edu­ca­tion, in her case from Albany State Teacher’s College back in 1957. She then went on to earn a Masters degree in math­e­mat­ics at the University of Michigan, and at that point IBM had the good sense to recruit her to join their team.

At IBM, Fran quick­ly became a pio­neer in the field of com­pil­ers and also in high-performance com­put­ing. She is wide­ly rec­og­nized for her fun­da­men­tal work in the the­o­ry of pro­gram opti­miza­tion and also for her ear­ly lead­er­ship of the Parallel Translation project.

Fran has won many sig­nif­i­cant awards for her work over the years. Among them she is an IBM fel­low, which is an extreme­ly pres­ti­gious award at IBM. And in 2007, the Association of Computing Machinery announced that Fran Allen was the recip­i­ent of the Turing Award, which is the high­est hon­or for a com­put­er sci­ence researcher in our field. And she is the first woman to win this award.

Fran has also dis­tin­guished her­self in ser­vice con­tri­bu­tions to our com­mu­ni­ty, and par­tic­u­lar­ly in sup­port­ing women. She spent many years as a men­tor through IBM’s men­tor pro­gram. In 2000, she her­self became the first recip­i­ent of the Fran Allen Women in Technology Mentoring Award.

Please join me in giv­ing a warm wel­come to Fran Allen, our ple­nary speak­er today.

Fran Allen: I’m so delight­ed to be here, and I assume the slides will come up. I have so much to tell you. We’ve got a lot of work to do this morn­ing. I’m going to make quite a few sug­ges­tions, and I’m going to tell you about my own career. [Tech comes up and shows Fran how to use a slide con­troller on podi­um.] I need some tech­nol­o­gy help.

I’m going to talk about real­ly what it means to me and has meant to me to win the Turing Award. A great deal. And also to com­ment on the fact that the times they are a‑changing, for com­put­er sci­ence and for the whole impact that com­put­ing has had on our field, and on every oth­er field and on peo­ple around the world. So as I said, we’ve got work to do this morn­ing. So let’s get with it, if I get the tech­nol­o­gy worked out here.

So, change is hap­pen­ing. And as every­one knows, last year I won the Turing Award, and I just thought I would point out that I am the first woman since the award was ini­ti­at­ed by ACM in 1966, and there have been a total of 54 peo­ple [who] have received it. Now, I am deeply hon­ored by that, but also deeply con­cerned. And I think we all should be very deeply con­cerned that more of the women in our field are not receiv­ing, in my opin­ion, the awards that they deserve.

Anyway, before I get into my rants on some of that, and I real­ly will try not to do a rant, I do want to talk about a num­ber of things. There’s a list [slide; not shown] of what I’m going to talk about. But since Anne[sp?] has giv­en such a nice intro­duc­tion about my back­ground before I joined IBM, I’m going to start with how I joined IBM. 

IBM had a recruit­ing brochure in 1957. [laugh­ter] And I did­n’t see it at the time when I was recruit­ed at the University of Michigan, fin­ish­ing up my Masters degree in math­e­mat­ics. But I came across this quite a few years lat­er, and I’m cer­tain, or it’s like­ly, that I was recruit­ed as a result of IBM’s expand­ing and its need­ing many peo­ple at that time, and look­ing for women. It’s always had a great his­to­ry of hir­ing women, and oth­er minori­ties, of course, going way way back.

And this was, I want to point out, well before com­put­er sci­ence. Computer sci­ence did­n’t exist when I start­ed in ​’57, and it did­n’t come into exis­tence until at least ten years lat­er. And I’m going to come back to that point at some stage.

So the first project after I joined IBM, since I’d been a high school math teacher for a while and joined as a pro­gram­mer, my man­ag­er decid­ed that I should teach the new FORTRAN lan­guage, which had come out a few months before I had joined. And they were going to issue an edict that every­body, all the sci­en­tists in IBM research were going to have to use this lan­guage and give up on using Assembly lan­guage. And they had to take a course. So I was the teacher of the course and I had to learn FORTRAN and teach this unhap­py class of sci­en­tists how to use it.

Well, the net of it was that I was enam­ored with the whole idea of using a high-level lan­guage and adopt­ed— And I did­n’t real­ize I’d done it at the time, but over the years I’ve real­ized I adopt­ed the goals that John Backus had set in estab­lish­ing FORTRAN, for user pro­duc­tiv­i­ty and prob­lem per­for­mance. So that became the goals of my entire career in high-performance com­put­ing and com­pil­ers and com­put­er lan­guages for high-performance computing.

Part 2

So there’s some won­der­ful sto­ries that go with that. But I want to move on quick­ly through this and talk about my next project, which was Stretch, again at IBM. You’re going to get some IBM his­to­ry here, some ear­ly his­to­ry that’s not very well-known. And in fact, because it isn’t very well known (cer­tain­ly the Stretch is not well-known for some rea­son) I’ll tell a lit­tle bit about that. The Computer History Museum two or three weeks ago had an event in which Fred Brooks and I… (Fred was asso­ci­at­ed with the Stretch project, and I was asso­ci­at­ed with the Stretch project) and one oth­er per­son, Harwood Kolsky did a sort of town hall meet­ing on that par­tic­u­lar project. 

Its goal, and this was set by the President of the com­pa­ny, was to be 100X faster than any­thing that exist­ed. Actually, the first one was built for Los Alamos. And com­put­ing at the time was still extreme­ly prim­i­tive, and com­put­ers were prim­i­tive. And ​“stretch” was absolute­ly the right word for it. 

I point out on the sec­ond line [slide not shown] a prob­lem that the design­ers rec­og­nized actu­al­ly in 1955, that mem­o­ry access time was going to be their biggest chal­lenge. It’s still our biggest chal­lenge in com­put­ing. We don’t know how to get the data to the com­pu­ta­tion, and it’s still… Caches, all kinds of things, since then and it’s still a huge, huge problem.

So the total project involved an extra­or­di­nary piece of hard­ware, and some extra­or­di­nary com­pil­er to go with it, an I was part of the com­pil­er activ­i­ty. But in addi­tion to the Stretch part there was anoth­er attach­ment, big­ger than Stretch, and a total­ly dif­fer­ent machine for the National Security Agency, and I was involved with that also. Not with the hard­ware at all, but with build­ing the com­pil­er for it and a lan­guage for it.

That still would be, out­side of base tech­nol­o­gy, an amaz­ing machine today. It was a machine host­ed by Stretch. It had a stream­ing data com­pu­ta­tion­al mod­el. It was one of a kind. It had eight instruc­tions with unbound­ed exe­cu­tion times. And of course the basic pur­pose of the machine was to do code-breaking on data col­lect­ed up on lis­ten­ing sta­tions around the world. And there was anoth­er part to it I don’t men­tion here where the data was stored, which was an inte­gral part of the com­pu­ta­tion­al units. And it was the only sys­tem that has ever been built that had perfectly-balanced I/O, mem­o­ry, and com­pu­ta­tion­al speeds. So hun­dreds of thou­sands of bytes of data could flow through and be ana­lyzed— Looking for pat­terns, of course that’s what code-breaking is large­ly about. Looking for pat­terns an accu­mu­lat­ing data as a result of the Harvest section.

One of the great design­ers of that, and Fred and Harwood and I agreed at Computer History Museum event that the great­est design­er on the whole sys­tem of Stretch and Harvest was Jim Pomerene, who is still alive. He came to IBM after work­ing with John von Neumann on John von Neumann’s com­put­er. He was the lead engi­neer there. And he has not been prop­er­ly rec­og­nized over the years. And in think­ing about Jim, I real­ize that there are many many oth­ers that have not been appro­pri­ate­ly rec­og­nized over the years for their work. Now, in his case a lot of his work was top secret. Some of it was very very ear­ly, but he still to this day is work­ing with peo­ple on var­i­ous prob­lems relat­ed to these kinds of systems.

My role was again the com­pil­er (I’ll show you a pic­ture of it in a minute) but also a lan­guage called the Alpha lan­guage for the code-breakers to use, and it was a per­fect match for this very special-purpose machine in that the code-breakers could write their code very suc­cinct­ly, their algo­rithms, and have it map to this very spe­cial machine in a very short few lines of code. I recent­ly looked to see how many lines of code would it take to do a DNA map­ping, if one had the sys­tem and had the Alpha lan­guage, and it was about fif­teen to do a total DNA mapping.

Now, we don’t have lan­guages like that any­more, or not many of them, that are so high-level and so very special-purpose. As we move for­ward from where we are and what our next steps have to be (and I’ll talk about that in a lit­tle bit) we’re going to have to be rethink­ing a lot of how we do com­put­ing these days. And we’re at a tip­ping point in the com­put­ing field, and a lot of changes are going to be hap­pen­ing. I’ll get to that. But going back to look­ing at some­thing like this, it’s clear we can rethink some of the lessons that were learned at that time, and in this par­tic­u­lar­ly bold project.

[slide not shown]

Here’s the com­pil­er we built then, and it’s still… The pat­tern of this com­pil­er being able to take mul­ti­ple source lan­guages, very dif­fer­ent source lan­guages. FORTRAN, a busi­ness lan­guage COBOL was pop­u­lar (It was sort of like COBOL.) and the Alpha lan­guage, which was the lan­guage that the agency want­ed. And then doing a for­mal analy­sis going down to an inter­me­di­ate form, going through opti­miza­tion and trans­for­ma­tions, and then on an inter­me­di­ate form that could rep­re­sent all of the source lan­guages, and then map­ping those to the two dif­fer­ent tar­get machines… IBM’s prod­uct com­pil­er today, the fam­i­ly of XL com­pil­ers, looks like this in some sense, and it can take all of the prod­uct lan­guages that IBM has for its cus­tomers to all of the machines that IBM has towards cus­tomers, though a sys­tem that looks sort of like this, in inter­me­di­ate forms.

So what were the out­comes of this project? Well, the Stretch did­n’t reach its 100X goal. (I’m hav­ing to look at my own slides because I thought I’d have my own com­put­er here. They took it away.) But the President of the com­pa­ny, Tom Watson, had announced a 100X goal on this project, and then he had to apol­o­gize to the world. And in fact the FORTRAN com­pil­er, which should’ve been the sim­plest part of it, also failed in terms of how for anoth­er cus­tomer, (this was for weath­er fore­cast­ing), it took 18 hours to do a 24-hour fore­cast. It was both the machine was not as fast as it was sup­posed to be, and the com­pil­er was not as fast—it did­n’t pro­duce as good code as it was sup­posed to. 

But we were invent­ing every­thing, and it was a group of very young peo­ple. And no expe­ri­ence. And there was­n’t any expe­ri­ence to be had any­way. We would­n’t have tak­en on this project, I think, if we knew what we did­n’t know.

Part 3

And that was true across the board on that project.

I might say, as far as the com­pil­er was con­cerned and the project itself, there were many women involved. Four of us were first-line man­agers on the com­pil­er, and three of us were women. And this is in [the] late 50s. But com­put­er sci­ence did­n’t exist yet. That’s not com­put­er sci­ence’s fault per se, but I did make a link lat­er on which had to do with how the pro­fes­sion­al­ism of the field evolved over the years as to why we’ve had some prob­lems with [that?].

So you see a com­ment there [slide not shown] by Dag Spicer from the Computer History Museum. They have one of the machines, the Stretch. The Harvest dis­ap­peared into Fort Meade. I was at Fort Meade for a year and actu­al­ly was there in the base­ment dur­ing the Cuban Missile Crisis. It was a mar­velous and very inter­est­ing, very scary, expe­ri­ence. And I thought I would be spend­ing the rest of my career there because I was sup­posed to write the accep­tance test code, and it was to do auto­mat­ic abstrac­tion of arti­cles, and I did­n’t. So I wrote up in Alpha an auto­mat­ic abstracter and I nev­er thought it’d be able to abstract Time Magazine arti­cles, but it did the sec­ond time, and I got to go home. But I was amazed about that.

Okay, just going quick­ly, the next project that I was involved with (though 360 came along in here, too) as far as IBM his­to­ry is con­cerned was to build…those of us that were dis­ap­point­ed we did­n’t real­ly suc­ceed on Stretch, that we were to build the fastest machine in the world, again. So we did that. Joined a project to do that, we did­n’t do it. It was cancelled.

And this time we had moved to California. We were in Silicon Valley, before Silicon Valley was there. And we moved there in order to get away from the Armonk head­quar­ters of IBM. They were dis­tract­ed by what was going with anoth­er big project, the 360 system.

So this, we got a long ways, and did build an exper­i­men­tal com­pil­er, because we did­n’t know what the machine would look like. So we built the exper­i­men­tal com­pil­er to be language-independent and machine-independent. It was a carry-forward of the very first piece of work we had. I had John Cocke’s pic­ture up there. He was an ear­li­er Turing Award win­ner, and this was his favorite project, and he was one of the dri­vers of the engi­neer­ing; just absolute­ly amaz­ing man. I’ve had the good luck all through my career of work­ing with some real­ly tru­ly great peo­ple, and that’s been very fortunate.

But there weren’t real­ly any women on that project. It was inter­est­ing that seemed to, in that peri­od of the 60s, start to fall off. 

When this col­lapsed in Silicon Valley every­body left to go every which way, I actu­al­ly stayed on for a while and tried to get the com­pil­er embed­ded in some work that Gene Amdahl was doing (I worked for Gene Amdahl, if you know him) but that did­n’t pan out.

So I took a sab­bat­i­cal at NYU, and my career has always had a cir­cle of research, going to prod­uct and, mov­ing often with the research out to prod­uct, and then often with doing a sab­bat­i­cal or some peri­od of time where I look again at what the research issues are. And going to prod­uct is a great way to find out what the research issues should be. You spend all your time not solv­ing inter­est­ing prob­lems, but try­ing to meet a dead­line, and stack up new inter­est­ing areas to explore.

Then I spent a peri­od kind of get­ting my things back togeth­er again, doing a lot of prod­uct work, com­pil­er work, adding on, writ­ing some papers. And then IBM, which was very late in get­ting into par­al­lelism, asked me to put togeth­er a com­pil­er group to look at auto­mat­ic par­al­lelism. And that turned to to be one of the most fun recent projects I had. I went to vis­it the University of Illinois, where great par­al­lel work was going on, and NYU and some oth­er places, and hired a lot of young peo­ple and we put togeth­er a team that for 10 or 15 years just poured out papers and com­pil­er tech­nolo­gies and built pieces of it and every­thing. I real­ly real­ly real­ly enjoyed that whole era.

Now, I skipped over what hap­pened when I first came back to research after this kind of round-trip to the West Coast, and I found a glass ceil­ing. It’s tak­en me quite a long time to under­stand why the envi­ron­ment had changed. And I am con­vinced, and every­one I speak to guess­es that it’s the same rea­son… What I believe is that com­put­er sci­ence emerged as a sci­ence, as a pro­fes­sion, with all the require­ments on what pro­fes­sion­al stan­dards and require­ments of what one need­ed to know to get a job in the field.

Most of the com­put­er sci­ence depart­ments in the mid-60s emerged out of the engi­neer­ing schools, which are almost all men. And the engi­neer­ing schools had cours­es that the stu­dents had to [take], and the com­pa­nies then were hir­ing peo­ple that sat­is­fied cri­te­ria that did­n’t exist ear­ly on when com­put­er sci­ence did­n’t exist. Earlier, when any­one who had good marks and [was] bright and eager, I guess…bright-eyed and bushy-tailed, could get a job. And they did­n’t have to be math­e­mati­cians or sci­en­tists. They could be English majors, or what­ev­er, because the field was grow­ing and no cre­den­tials had been established.

In that peri­od, then, cre­den­tials were estab­lished, and by the ear­ly 70s things had real­ly changed for women, at least in my envi­ron­ment, and most oth­er groups that I’ve talked to about this the­o­ry absolute­ly agree that that was where there was a sig­nif­i­cant shift.

Okay, mov­ing forward. 

I retired, by the way. I did­n’t men­tion that. I did it after 45 years. And then after 50 years of being asso­ci­at­ed with the field, I received the Turing Award and my life changed. [applause]

Part 4

And I want to tell you what it was like to get a phone call from Ruzena Bajcsy, actu­al­ly, she was chair of the Turing Award committee—this was in January or February of 2007—to tell me that I was receiv­ing the award, which was stun­ning. And then I of course got calls from every oth­er per­son on the com­mit­tee. But I was warned that I could­n’t talk about it for two or three months. ACM and IBM need­ed to get the PR togeth­er. And I was kind of hap­py that there was a delay in that time, because I had time to think about what this meant to me, and in the first hour of hear­ing the news… I was home alone and I just kind of paced around and talked to my cat. And real­ized that one of the things… There were two things that came [up]. What in the world am I going to talk about? Because Turing Award peo­ple and Jim Gray, who was a for­mer win­ner and he talked with me that week. That fol­low­ing week­end was the week­end that he dis­ap­peared, very sad­ly. But he said, ​“Well, you’re going to be on every­body’s A list, so you bet­ter get start­ed think­ing about what you’re going to say.”

But the oth­er thing that I kept com­ing back to was the fact that so many women that I knew, that I’d worked with, and were deserv­ing of so much more recog­ni­tion than they had ever got­ten in their careers… And that’s true of men too, of course, like Jim Pomerene. But it real­ly made me feel that part of what I want­ed to do was to try and change that, change the way we rec­og­nize peo­ple and increase the recog­ni­tion. I now am on a com­mit­tee in Anita Borg that Katy Dickinson heads, try­ing to focus on awards for women. I spend quite a lot of time not so much on that com­mit­tee but think­ing about oh, we’ve got to nom­i­nate so-and-so, do some­thing about this, and stim­u­late those awards. And I think that, hope, just like men­tor­ing this is an extreme­ly impor­tant task for all of us. And it’s for both men and women. Equity is impor­tant here. But I think there’s been some equi­ty lack­ing for some people.

So what I’m going to do now is talk about, and I’ll be quick about the next sub­ject, is what did I decide I was going to talk about. I have a lit­tle some­thing to say about whith­er com­put­er sci­ence, and the last one is where of course have all the women gone? Well…they’re here. And that is great. This is just a real­ly uplift­ing feeling.

So I’m going to go very quick­ly through an excit­ing new prob­lem, which is both an oppor­tu­ni­ty for a lot of peo­ple, and a lot of women have got­ten involved with it. And it’s new. Well, let me just tell you what it is.

The per­for­mance of com­put­ing is not going to stay on the curve that is shown up there. Each dot on there is the fastest com­put­er of that year, start­ing with ENIAC way down in the cor­ner, 1944 I guess, and going up to the Blue Gene. But there’s that curve, which is Moore’s Law, which has been very steady and doing well in the high-performance field. But what is hap­pen­ing is that the dri­vers of this, the basic core tech­nolo­gies, are in trou­ble. This is not my spe­cial­ty, but some peo­ple here would know a lot more about it. But it’s in trouble.

The per­for­mance futures have fall­en off, and fell off about 2002 or 2003. I was talk­ing with Greg Papadopoulos this morn­ing and try­ing to pin him down on just when it had hap­pened. Anywho, close enough. So that is cre­at­ing a seri­ous gap in the expec­ta­tions of this area. And also there are prob­lems with heat and ener­gy. That’s a sep­a­rate prob­lem but due to the same root cause of miniaturization.

So now what’s the solu­tion? Well, the solu­tion is to move to have explic­it par­al­lelism, back to a field I know a bit about, and do it in soft­ware. And have what’s on the chip, the actu­al technology…the parts [can] be much sim­pler, much small­er, much cool­er. And put all the work that they used to do in terms of hav­ing capa­bil­i­ties on each of those chips, lots of heavy-duty capa­bil­i­ties in the tran­sis­tors on these chips, onto the code that’s pro­duced by software.

What they’ve done is giv­en us a tool which, with a lot of par­al­lelism, but it’s up to us in soft­ware to fig­ure how to use it. And this is a com­ment on the par­al­lelism, and in fact, we can­not, with­out a huge amount more work, we’re not in a posi­tion to either in soft­ware or to ask the users to do it. The archi­tec­tures upon which all these gad­gets are based and which are the tar­get of all the soft­ware that runs, is chang­ing as rapid­ly as I have ever, ever seen it, and it’s just get­ting start­ed. This is going to affect every­thing from the very high-performance com­put­ing, maybe even less up there, but it’s going to effect the hand­helds and everything.

So we’re at a fair­ly frag­ile crux point in com­put­ing [for] basic assump­tions we’ve been mak­ing about com­put­ing for a long time, for 60 years. And the soft­ware’s not in a posi­tion to help, to do it, and nei­ther can the users grap­ple with all of these myr­i­ad archi­tec­tures that are going to emerge at a very steady pace.

Part 5

So John Hennessy, the President of Stanford and of the Hennessy-Patterson books has said that it’s the biggest com­put­er sci­ence has ever faced, and I absolute­ly agree with him. I’m not sure that there’s any dis­agree­ment on that state­ment. I see it as also the biggest oppor­tu­ni­ty we’re going to get to make some very nice advances in com­pil­ers and lan­guages and par­al­lelism. But it’s going to mean new lan­guages, new com­pil­ers, new sys­tems, and it’s not going to hap­pen overnight. Fortunately the com­mu­ni­ty rec­og­nizes all of that, and the com­mu­ni­ty I think itself, from var­i­ous com­pa­nies and cer­tain­ly uni­ver­si­ties, are all com­ing togeth­er and coop­er­at­ing on dif­fer­ent ideas and aspects of this issue. It’s real­ly a time of change for our field, and a time when I think since there are already quite a few women work­ing on aspects of this prob­lem, that it’s going to be an oppor­tu­ni­ty for the women to play major roles. Partly because they’re such good community-builders and com­mu­ni­ca­tors. I won’t go down that…why, but it real­ly is hap­pen­ing in every cor­ner of our field.

As I said, the change is hap­pen­ing. This is a lit­tle side­bar. I real­ly have been unhap­py about com­put­er sci­ence, but I’m real­ly not in the mid­dle of the issues of edu­ca­tion in com­put­er sci­ence, of the cur­ricu­lum. But it dis­tress­es me that I see less and less inter­est­ing work being done in my own field. Piles of papers and con­fer­ence after con­fer­ence, and I don’t find much new hap­pen­ing. So I’ve been dis­tressed about the field itself, and I heard Dick Carp, who’s one of the great the­o­rists in the field (also a Turing Award win­ner; he’s at Berkeley) say in a talk a cou­ple of weeks ago that com­put­er sci­ence ​“is plac­ing itself at the cen­ter of sci­en­tif­ic dis­course and exchange of ideas, and this is only the begin­ning.” He also says the algo­rith­mic world[view], which is part of our com­put­er sci­ence is chang­ing math­e­mat­i­cal, nat­ur­al, social, and life sci­ences. So I think that this is absolute­ly great news from one of our great thinkers. It means change is hap­pen­ing in a very fun­da­men­tal way.

Switching now to more of where we’re going in wom­en’s issues. This is first of all my hopes for the future. I’d like to see a new gen­er­a­tion of women expe­ri­ence the excite­ment I feel for the whole of my field. I would real­ly like women to stay in tech­nol­o­gy, stay in the sci­ence side. It’s ter­rif­ic when they become great lead­ers, but not all women are des­tined to do that. It’s great when you can, but there’s a hap­pi­ness, a joy that I find every day, in the field that I spe­cial­ized in for the last 50 years. I get excit­ed by new ideas and the new oppor­tu­ni­ties. And dis­cour­aged some­times, of course. But it’s a great feel­ing to be part— And it’ll go on the rest of my life, I’m sure. As long as I have my mind. One always starts to worry.

And I’d like to see women cre­at­ing the work­place that meets their indi­vid­ual needs. We’re at a point where we have so much won­der­ful tech­nol­o­gy for com­mu­ni­cat­ing and shar­ing, and it’s glob­al. And a glob­al com­pa­ny like IBM has projects where the indi­vid­u­als or peo­ple on the projects work on com­po­nents of the project around the world. If there’s an expert in Bangladesh, that per­son is part of the project and so forth. And well, you know the whole glob­al­iza­tion issue, it’s adding a great deal to the diver­si­ty of our ideas and to the ben­e­fit that we derive from it.

And the third one is I’d like to see com­put­er sci­ence become a core sci­ence. This was inspired by Dick’s remarks at the British Computer Society work­shop on visions in com­put­ing last week.

And I’d also like to see Anita’s goal hap­pen. She had a goal for us to have 50% women in the work­place by 2020. Anita had been a stu­dent of mine when I had a com­pil­er course at NYU, and I was the only woman [and] was a pro­fes­sor of hers in grad­u­ate school. We kept up a rela­tion­ship over the years, and I remem­ber one very rainy night I was at work late and I was in a ter­ri­bly grumpy mood about a lot of things. The phone rings and it’s Anita, and she said, ​“What do you think about 50/50 in 2020?”

I said, ​“Oh Anita, we can’t pos­si­bly do it. It’s 2000 something-or-other. It’s impos­si­ble, it’s impos­si­ble. Look at what’s going on in the grad­u­ate schools and every­where. We can’t do that.” So I real­ly poured cold water on it. 

And then she said, ​“Oh, well.” She said goodbye. 

So I real­ly start­ed to think about it, and about half an hour lat­er I called her up and I said, ​“Yes.”

She was great for putting out grand chal­lenges for all of us, even those of us who would get cold feet once in a while.

And [I] cer­tain­ly want to have many many more women, and men, but espe­cial­ly women, have the expe­ri­ence of win­ning an award as I have done.

And I’d like to spend just a lit­tle bit on remem­ber­ing a few great women. There are many many many many, and I would be in real trou­ble if I tried to make a list. But Anita is cer­tain­ly one of them.

Part 6

When I was in England a cou­ple weeks ago, I vis­it­ed Bletchley Park, which was the code-breaking orga­ni­za­tion or site for the British dur­ing World War II. It’s a muse­um that’s in des­per­ate need of fund­ing, and for a while I’ve been a long-distance mem­ber of the muse­um because of my NSA inter­est. But I had not ever vis­it­ed it. Go there, if you pos­si­bly can. And look at their web site. This is where Turing worked and lived and did mar­velous math­e­mat­ics and insights.

Also, I dis­cov­ered, there were thou­sands of women then ​“wrens,” they were called) who were in the ser­vice and had been draft­ed to mon­i­tor the German com­mu­ni­ca­tions. Of course there was the Enigma, if you’ve heard of that machine. It was one of the German machines, and the British had their own.

Then there’s the ENIAC com­put­er, which was being built here in the United States. And by the way, the one at Bletchley Park, though it’s been secret for a long time, was the first stored-program machine. ENIAC came a year lat­er. So tech­ni­cal­ly it was not first, but it does­n’t mat­ter. It was essen­tial­ly built kin­da simultaneously.

Jean Bartik, who was one of the com­put­ers on that project, is being rec­og­nized by the Computer History Museum next week as a fel­low of the Computer History Museum. She’s still alive and feisty. The pro­gram­mers were called ​“com­put­ers” at that time, or on that project at least.

I’d also like to think, as I did when I received the award, about all the many peo­ple… One was par­tic­u­lar­ly a friend, Betty McDonough, who was on the ear­ly days of Stretch and built the very first multi-tasking inter­face machine as part of the mem­o­ry laten­cy prob­lem, solv­ing it. And the work was just stolen from her. Terrible. But that’s the way it happens.

And this is the pic­ture of the Wrens work­ing on the Colossus machine. That’s one of the machines that they had at Bletchley Park. For years, three shifts a day, these peo­ple would work and nev­er was the infor­ma­tion leaked. It was just amaz­ing. Unbelievable.

I want to thank all of you. And I would like as my final word, I guess, is that wom­en’s work should not be top secret like the Wrens’ work and all of Jim Pomerene’s work. And we’re ready to cel­e­brate. We have a rea­son to cel­e­brate. Because we have more and more women, great women, enter­ing our field now. I don’t know where they’re com­ing from, because we do have a prob­lem in com­put­er sci­ence and com­put­er sci­ence is chang­ing, but I real­ly am excit­ed about these times, and it’s time for a celebration.

And here is a great pic­ture of a water­col­or that Maria Klawe did while she was attend­ing a meeting. 

So thank you very much, and good luck.

Introducer: Thanks, Fran. That was a real­ly insight­ful and fas­ci­nat­ing look back to the pre-history of com­put­er sci­ence all the way to future, both on the tech­ni­cal and to some degree the cul­tur­al front. I par­tic­u­lar­ly enjoy the way you look at the per­for­mance chal­lenge as an oppor­tu­ni­ty rather than a prob­lem, and I think we’re all look­ing at the issues of rep­re­sen­ta­tion of women in com­put­er sci­ence in exact­ly the same way.

So every­body, let’s thank Fran again, and enjoy your cof­fee break.