Fran Allen, IBM Fellow Emerita and 2006 Turing Award Winner
From Anita Borg Institute Wiki
Fran had a wonderful introduction. She started her career with a degree in education at the Albany state teachers college in 1957. She later received a MS in mathematics at the University of Michigan, and an IBM recruiter “had the good sense to” hire her. She is a pioneer in compilers and high performance computing, is an IBM fellow, and was the first woman to win the ACM Turing Award.
Fran's first slide included the quote "The times they are a-changin’ " from Bob Dylan 1964. Another slide stated that “computing is a tipping point and women are the change agents of the future," which is something that she truly believes in. There are currently 54 Turing award winners, and although she is honored to be the first woman to receive the award she is also concerned that she is the first and only woman winner. As she expounds on later in the talk, she strongly believes that too many women and (and even some men) do not receive the credit or the awards that they deserve.
She showed a picture of the IBM 1957 recruiting brochure that said “My Fair Ladies” and had flowers on it with an “IBM” tag on the flowers. This is the year she joined IBM as a programmer. Her first project at IBM was to teach the new FORTRAN language to everyone in IBM research, as they would be required to use it instead of assembly language. Her teaching background is one of the key reasons she was chosen for this assignment. Ironically, the goals of her career became user productivity, and program performance, which are also Backus's goals with FORTRAN! The next project she worked on was the Stretch project (1956-1961), that had a goal of creating a machine that is 100 times faster than existing machines. They determined that memory access time was a main performance limitation, which is still true today. Next was the Harvest project with NSA (1958-1962), hosted by Stretch. The programming language ALPHA was designed specifically for their code breaking problem and this machine. The machine did code breaking on data collected from around the world for national security. It is the only system that has ever been built with perfectly balanced I/O, memory, and computational speeds. There are many people who have not been recognized for their work on this project, even though they should be.
We need to be rethinking about how we do computing, as we’re at a tipping point in the computing field and a lot of changes are going to be happening soon. Looking back at these old projects, we need to rethink the lessons learned in that time with bold projects that really pushed the bounds of computing. The compiler for the stretch-harvest project took 3 languages and combined them together, and mapped to the two different target machines; some IBM compilers today still look similar. For the Stretch-Harvest project, they didn’t get 100 times better goal, only got half of that. Although that technically made it a failure, it still was a great project. They had to create everything for it themselves. There were many women on this project, 3 of 4 first line managers were women! But of course, CS didn’t exist yet at that time.
They started a new project called Advanced computing system (ACS), whose goal was to build the fastest computer in the world. The project was canceled though, because the project was too costly and too big. A quote from the slide at this point is that “people scattered...but the compiler lived on!” No women were really on that project though. It seems to her that the 60s seems to be when women started to fall off from this type of work. After that project she wrote papers, took a sabbatical at NYU, then went back to IBM research and hit a glass ceiling. She went from going to product, to learning about current research over the course of her career, always moving back and forth. She believes that going to product is a great way to find out what research problems SHOULD be investigated! She was later asked to put together a compiler group at IBM to look at parallelism: these were the most fun recent projects she had, she got to visit universities, hired young people, got a team together that for 10-15 years just poured out papers and compiler technologies!
It has taken her a long time to understand why the environment at IBM had changed when she got back such that there WAS a glass ceiling for her to hit. She believes that as CS emerged as a science and a profession with all of the requirements of professional standards and what you needed to know to get a job we lost women, because most of the CS departments in the mid 60s came from engineering schools that were almost entirely men. Companies then were hiring based on criteria that didn’t exist when the field unofficially existed before that time. In that period, by early 70s things had really changed for women in her environment and it seems from talking to friends in other computing areas that it was true for them as well.
When she found out that she had won the Turing Award, there was a delay of a few months before she could tell anyone. When she first found out she wandered around her house talking to her cat about it, but she is glad to have had time to think about what it meant to her to win the award. She mostly wanted to figure out what to talk about at the award ceremony, and the advice she was given by a previous award winner was to start thinking about it ASAP! As she thought about it, she kept coming back to the fact that she knew so many women who were deserving of so much more recognition than they ever got in the careers. Part of what she realized she wanted to do was to try to change that problem, so that we recognize people is a way that works better and ensures more deserving people are awarded. She is now on committee with the Anita Borg Institute to work on awards for women. She believes that just like mentoring this is extremely important, as making sure that we have equity in awards will benefit everyone.
There is an exciting new problem she wants everyone to think about related to the fact that the fastest computer of the year tends to double every 1.5 years. We can’t keep up this curve forever, so we must find a new solution. The drivers and core technology are in trouble. Real performance of transistors fell off the expected growth curve around 2002. There are heat and energy problems as well, due to the same miniaturization. The solution to this problem is multicores! She is a proponent of explicit parallelization in software instead of hardware, so that what is on the chip can be simpler, easing some of the barriers we are facing with increasing computing ability. Current software cannot provide the parallelism needed for this new challenge. Users can’t do it either. Hennessy says that this challenge is “the biggest problem Computer Science has ever faced.” Women can and should be involved with this challenge.
It distresses her that she sees less and less interesting work being done in her own field in research. She feels that not much new seems to happen anymore. She’s not involved in the education of students at universities, but is interested in big problems being solved again.
She has specific hopes for the future. She wants to see a new generation of women that experience the excitement she has for the field. She wants women to stay on the science side, or become great leaders (although not everyone is meant for or interested in that). She is still excited by new ideas and new opportunities and is sure she will feel this way the rest of her life; it’s a great feeling to be a part of this community. She would also like to see women creating the workplace of their individual needs. There’s great technology for communication and sharing, we work with people around the world, and we should be able to leverage this to make the type of workplace we desire. She would like Computer Science to become a core science, especially one that is of more interest to women.
She would love to see Anita’s goal happen: the field to be 50-50 by 2020! Anita called asked her if she thought this idea was possible one evening when Fran was in a bad mood on a rainy evening, and she said there’s no way it could happen! They hung up the phone, but Fran starting thinking about the suggestion. About 30 min later she called her back and said “yes.” On a related note, she would certainly also like to have more women Turing award winners. To close, she remembered a few of the many great women: Anita Borg, the Wrens at Bletchley park (code breaking organization for the British during WWII), ENIAC “computers” such as Jean Batik who is being recognized by the computer history museum next week (and “is still alive and feisty!”), and Betty McDonough who invented multi-programming and never got credit for it. There were 1000s of women called “wrens” at Bletchley park who were drafted to monitor the German communications during WWII, and it is amazing how nothing was ever leaked from their work. The machine used at Bletchley park was technically the first stored program machine, as it beat the ENIAC by about a year.
In the end, she is ready to celebrate, and believes we have reason to celebrate. There are many great women entering the field now, and she is very excited about these times.
