At Illinois, Professor Laxmikant Kale Found An Enduring Path In Research And Education
As Laxmikant “Sanjay” Kale recalls it, 1985 was a good year to interview for jobs in computer science.
There was no doubt, he said, about which one was the best fit.
“Illinois was the one that stood out because of its friendly culture,” Kale said. “I didn’t sense any empires clashing with each other or politics, and it was one of the top-flight schools. It was an easy choice.”
Kale retired from Illinois CS in June after 34 years, leaving as the Paul and Cynthia Saylor Professor of Computer Science.
Kale has devoted much of his career to parallel computing, working to enhance performance and productivity. He is known for his pioneering ideas of parallel programming with an adaptive runtime system, based on migratable objects and message-driven execution. These led to successes in automating such tasks as dynamic load balancing, communication optimization, and fault tolerance.
These ideas were embodied in Charm++, a parallel programming system that his Parallel Programming Laboratory has developed over the past two decades, which has been used in a multitude of collaborative parallel application codes in science and engineering.
Those applications included the simulation program Nanoscale Molecular Dynamics, or NAMD, created with the late University of Illinois Physics Professor Klaus Schulten. Built on Charm++, NAMD was designed for high-performance simulation of large biomolecular systems, a breakthrough that was rewarded with both the ACM Gordon Bell Prize and IEEE Computer Society Sidney Fernbach Award. For his contributions to parallel programming techniques, Kale was also named a fellow of the ACM and IEEE.
Kale explained that when he and his wife, Lata, made the decision to come to Illinois, they didn’t believe they’d stay long term.
“We were not consciously thinking that we would be here the rest of our career,” Kale said. “Champaign-Urbana is a place that grows on you. It’s hard to leave.”
Again and again, he would find, there were just too many reasons to come, and then to stay.
First was that friendly, collaborative culture of the town, campus, and department.
Then there was something else that first drew and then helped keep Kale at Illinois: The department was home to some of the true leaders in computer science, starting with the department head at the time, Bill Gear.
“Bill Gear was not only a fixture of the department — even in India, whatever I learned about initial computer design, I learned from his book,” Kale said.
Ed Reingold, another computer scientist whose book on algorithms Kale had studied, was also part of the department. And the pioneering C. L. David Liu was here, too.
“It was like a dream, coming to work among legends,” Kale said. “That was just amazing — you get used to anything, but that was amazing.”
And then, in 1991, Kale’s career took a turn that linked his early research, all the way back to his work as a student on logic programming and how to do that in parallel, to the path that would eventually lead to Charm++ — not to mention creating the most productive collaboration of his career.
“I’d just gotten my tenure, I was going on my first sabbatical, and Klaus Schulten came to my office and suggested we collaborate,” Kale said.
Schulten would eventually become a leader in computational biophysics, known for, among other things, his contributions to computational biology. But at the time he’d only been at Illinois a few years.
“So we started working together, looking at molecular dynamics,” Kale said. “I had started working on a parallel programming system that we then called Charm.”
The work with Schulten gave Kale something he’d longed for: A project that was application-oriented and linked that practical side of parallel computing with a computer science-centered penchant for abstractions.
NAMD would go on to be the Gordon Bell winner in 2002, and Kale and Schulte shared the 2012 Fernbach Award for their work.
In 2013, NAMD was used in the first simulation of the chemical structure of the HIV capsid.
And Charm++ continues to be influential, used, for example, in the collaboration between Kale and Professor Tom Quinn at the University of Washington on computational astronomy. And in May, Kale hosted the 17th Annual Workshop on Charm++ and its Applications, drawing researchers from around the country.
But Kale also takes great pride in his research group’s success in distributing and supporting software embodying his research ideas, including Adaptive MPI and the BigSim framework. He and his team won the HPC Challenge award at Supercomputing 2011, for an entry based on Charm++.
Kale graduated about 40 PhD students and close to 50 MS students in his 34 years at Illinois Computer Science.
He says one of the things he enjoyed most is that interaction with students.
“I feel satisfaction that the students that I had were always fired up by the novel approach we were developing,” he said. “I (will) miss regular teaching — I enjoy standing in front of a class and lecturing, giving a concept, seeing a concept take root in a student’s mind and seeing it grow or standing in front of a whiteboard and discussing research ideas with students.”
Before coming to Illinois, Kale received a B.Tech degree in Electronics Engineering from Benares Hindu University, Varanasi, India in 1977, and an M.E. degree in Computer Science from Indian Institute of Science in Bangalore, India, in 1979. He also worked as a scientist at the Tata Institute of Fundamental Research from 1979 to 1981.