People of ACM European Chapters - Alexandra Jimborean
December 5, 2017
Can you briefly describe your own line of research and how you became interested in this area?
I conduct research on software/hardware co-designs, with a focus on compiler support for shifting hardware complexity in software to achieve new levels of energy efficiency without tradeoffs in performance. Our world is becoming increasingly digitized and thus computing systems are prevalent in all aspects of our lives—from the laptops, computers and handheld devices that we operate daily, to the large server farms processing our online queries and interactions, to the plethora of devices known as the Internet of Things present in modern cars, smart houses and cities, wearable technologies, etc. Until recently, performance was the leading force in hardware design and the utmost software optimization goal. However, these new hardware trends impose a new constraint: energy efficiency. Energy has become the limiting factor in computing systems, with power and cooling costs limiting server performance, and battery runtime limiting mobile and embedded systems performance. In this new landscape, compilers play an essential role in achieving high performance and energy efficiency by matching the software to the more restricted hardware capabilities.
Understanding software behavior on such heterogeneous hardware, developing automatic optimizations for performance and energy efficiency and generating code that does not only gel on the underlying computer architecture, but also guides the development of emerging architectures, has challenged and motivated me. We have approached this problem from several ends. First, we proposed compiler techniques to transform software and successfully reduced energy expenditure when executing on complex, high performance hardware, without harming performance. Second, we targeted energy-efficient, simple architectures and delivered higher performance by providing compiler support to compensate for the limited hardware. And third, we proposed co-designed solutions in which software and hardware share responsibilities to strike the right balance between performance and energy efficiency.
In your area of research, what recent advance/emerging subfield will yield important advances in the years ahead?
Software/hardware co-designs have the potential to become the norm in new-generation platforms, as software provides the required flexibility and adaptability at low energy cost, while hardware provides the runtime support that translates to pure performance. As recent contributions show that co-design is the way forward, such automatic tools will become mainstream and will extend to include all other system layers. Furthermore, the dual-optimization goal we face today—performance and energy-efficiency—will expand toward a multi-objective perspective that may add portability, reliability, scalability, predictability, productivity, and so on, as new optimization goals. It will therefore become more and more challenging for programmers to write optimized code; hence co-designed automatic tools will be key to releasing the programmer from the burden of optimizing the code to meet all of these optimization objectives.
Given the rapid technological advancements and the fascinating spread of technology to all domains (scientific, medical, economic, etc.), researchers are being challenged to rethink standard software and hardware designs in order to keep up with new demands and unexpected optimization goals.
Can you tell us a little about your chapter?
The ACM-W Chapter at Uppsala University (UU), is still young and small, but very active. UU ACM-W was established in 2014 and it currently has 11 members, but many more collaborators, including many of the faculty members/staff, as well as the Gender Equality group from the Faculty of Science and Technology. The chapter is run by students. It promotes diversity and gender equality and is engaged in organizing activities that blend technical talks with career mentoring and networking. We seek to raise awareness of the importance of gender diversity in the computing world and offer an environment for social and professional interactions. The ACM-W UU Chapter not only promotes gender equality, but also demonstrates it through a balanced number of both women and men computer scientists that actively participate in the regular meetings and in the organization of the chapter’s events.
The UU ACM-W Chapter is a friendly group of graduate students that offer younger students the opportunity to meet regularly for fika to share and propose ideas, to ask for advice and help or simply socialize with their peers. (Fika is Swedish for a coffee break that’s more about socializing than drinking coffee. Swedes prefer not to translate the word. They don’t want it to lose significance and become a mere coffee break.) Technical talks as well as inspiring talks given by women in computer science are typically a success. Our panel discussions on gender equality and computing have also been popular with students. Among other activities, the UU ACM-W chapter organizes lecture series, coding events and hackathons with mentors from industry and academia. These events encourage students to participate, to go out of their comfort zone, and to tackle real problems under the guidance of their mentors. One of our most recent events, The Ada Lovelace Celebration of Women in Computer Science, gathered more than 80 participants from Sweden and around Europe. The celebration spanned two days, and included eight talks by representatives from industry and academia on technical topics as well as practical career advice. A hackathon gave students a chance to meet and program together with their mentors. The event was truly a success, attended with interest both by faculty and students of all levels and genders. As Faculty Sponsor, I am glad to see so much support from university personnel and local industry, and I am confident that the UU ACM-W Chapter will continue to grow.
What advice would you give to a younger colleague just starting out in your field?
The world of research is wonderful, as it nurtures creativity and dedication. If one believes in their goal, there are no limits to what can be achieved, created, discovered. There will always be some ups and downs, so one has to endure the challenges of assiduous work. I would advise young people to build a group of peers which offer support and help when needed and to find mentors that provide guidance and advice. Finally, I would also encourage young people to work hard to achieve what they believe in! Go off the beaten path, do not be afraid to learn and experience novelty, ask, be curious, and above all, be perseverant!
Alexandra Jimborean is an Associate Senior Lecturer in the Department of Information Technology at Uppsala University, Sweden. Her research interests include compile time and runtime code analysis and transformation, optimizations for performance and energy efficiency, and software/hardware co-designs.
Jimborean was awarded the Anita Borg Memorial Scholarship provided by Google in recognition of excellent research, along with 30 other distinctions, awards and grants. She was recognized with a Best Paper Award at the International Compiler Construction Conference (CC 2016) and a Best Presentation Award at the International Symposium on Code Generation and Optimization (CGO 2014). She serves as the Faculty Advisor for the ACM-W Chapter at Uppsala University.