Weaving Together Computational Power
How does one solve an intricate problem when available computing resources are not enough? One looks to grid computing, which can collectively use the power of hundreds or thousands of computers to work together to seek the answer.
Imagine an intricately woven fabric with strong threads interlaced to create a vast pattern, saturated with color and texture. Next, imagine that fabric stretching across the planet, conveying an infinite number of stories. Finally, imagine the loom on which the fabric was woven and the skilled craftspeople who meticulously threaded it in anticipation of the weaving.
Michael Lewis is one of those craftspeople. An associate professor of computer science in Binghamton’s Thomas J. Watson School of Engineering and Applied Science, Lewis was one of the original designers and architects of the influential Legion Grid Computing System while in pursuit of his PhD at the University of Virginia. “It was a ‘right place at the right time’ situation,” he says. “In the mid-90s grid computing wasn’t even a recognized field, now my advisor is best known as one of its pioneers.” Lewis and his team at the Watson School’s Grid Computing Research Laboratory (GCRL) sit at a virtual loom, creating and manipulating a vast technological fabric of algorithms and software services that will allow grids to stretch farther and wider.
Grid computing puts the computational power of hundreds, thousands, even millions of geographically and organizationally diverse resources at the fingertips of individual programmers, enabling them to solve problems that otherwise would be intractable by a single machine or a local area distributed system. Grid software generally runs on top of underlying host operating systems. It serves to manage the resources and make them available to applications developers. Researchers who could use this technology include geologists simulating wave propagation to predict areas of high shake damage and computational chemists modeling the properties of marine enzymes.
At the GCRL Lewis, his colleagues and students work on a host of projects including component-based grid application development, evolution of grid services and objects, high-performance C++ Web services tools, sensors and instruments on grids, and dynamic grid information dissemination.
For Lewis, an academic career in computer science has been about more than just expanding the field and its potential. “The research is great, but I also love the teaching and mentoring side of what I do. It enables me to encourage others, especially people from classically underrepresented groups, to pursue computer science as a path through life that can be interesting and rewarding,” he says. To that end, Lewis directs the National Science Foundation’s Research Experience for Undergraduates (REU) summer program, crafting a host of learning opportunities for students from other campuses, some of whom may not have access to graduate-level courses and faculty. As for the intricate fabric of his personal life, Lewis says he intentionally weaves in a great deal of time with his wife and three children.