With a peak speed of 360 teraflops, IBM’s Blue Gene is among the world’s fastest supercomputers.
Maria Eleftheriou (a research staff member at the IBM T.J. Watson Research Center) contributes to the design and implementation of parallel algorithms and parallel programming models for the supercomputer. She also performs large-scale simulations on Blue Gene to address questions of biological interest, particularly in the area of protein misfolding.
Eleftheriou began taking advanced courses in high school and has been involved in science and math ever since she can remember. In 1995, she received her master’s degree in engineering from Brown University. In 1999, she earned a PhD in theoretical/computational chemistry, also from Brown. After graduating, she worked briefly as a postdoctoral fellow at Columbia University in the BioMolecular simulation center.
In addition to her work with Blue Gene, she currently chairs IBM Research’s community of technical women called the Watson Women’s Network. The group serves as a vehicle for organizing seminars on technical topics, conducting round table events with technical leaders and executives, organizing and hosting social-networking events, providing professional mentoring, and discussing work-life balance.
The Anita Borg Institute (ABI) talked with Eleftheriou about her career, her advice for technical women, and about ways she keeps inspired and motivated in her line of work.
ABI: What has been the biggest challenge of your career?
Maria Eleftheriou: When I joined IBM, I had a mentor who kept telling me that the best strategy is to get out of your comfort zone; this positions you for greater success, but also greater failure.
The biggest challenge was to learn that it’s okay to fail, as long as you fail fast, cut your losses, and move quickly to the next challenge. This took me awhile to learn.
ABI: What advice would you give to girls who are thinking of going into a computer science or engineering field?
ME: My first and most important piece of advice is: start early with math and science to get a competitive advantage. By the time you go to college, it might be too late.
ABI: What keeps you engaged in your work on the Blue Gene project?
ME: Challenges and innovative projects motivate me. I enjoy working with parallel computing projects and developing algorithms for Blue Gene. I’m also involved in writing programs to use the Blue Gene machine. One of my contributions in parallel computing was to scale the 3D fast fourier transform (FFT) algorithms on thousands of processors. (Small data size 3D FFTs have poor scaling behavior on massively distributed machines.)
The other area I work in is scientific, exploratory, and computational biology. In collaboration with colleagues at IBM, I started a large-scale computation to better understand how the abnormal folding of the protein lysozyme (an enzyme found in egg whites, tears, saliva, and other secretions) is related to diseases such as Alzheimer’s and Mad Cow. We use the Blue Gene supercomputer to understand and compute how the lysozyme protein misfolds.
ABI: Do you have a particular mantra or saying that you turn to when things get tough?
ME: My current mantra that I heard recently from Randy Paush’s recorded final lecture, a professor of computer science at CS Carnegie Mellon University, is, “Brick walls are there for a reason — they let us prove how badly we want things.”
ABI: Do you have advice for women who have great ideas, but are unsure of how to make them happen?
ME: If you are unable to communicate your ideas in two sentences and explain their impact to the scientific and/or business communities, they will not survive. You are the best advocate for your ideas. Mentors and managers can provide advice and guidance and help facilitate your ideas, but you need to be highly dependent on yourself to make the ideas happen.
ABI: What about work-life balance?
ME: There is really no balance, it’s mostly work. But if you like your work, it’s really part of your life.