To pursue new research and application possibilities for a promising new computational technology – automata computing – the University of Virginia has established a new Center for Automata Computing, with seed funding from, and close collaboration with, the technology’s inventor, Micron Technology Inc.
Micron Technology, one of the world’s leading providers of advanced semiconductor solutions, Monday also announced the development of a fundamentally new computer architecture based on massively parallel automata technology.
“Micron recognizes the important role that research institutions can play in bringing this advanced technology to market, and we are excited to be working with the University of Virginia to establish the world’s first research center dedicated to the concept of automata-based computing,” said Paul Dlugosch, director of the Automata Processor Program at Micron Technology. “U.Va. will bring a strong multi-disciplinary approach to key research initiatives at the center. The University’s strong relationships and close proximity to the technology industry in Northern Virginia make U.Va. an ideal location to establish the center.”
Automata computing is based on simple computational elements that activate or deactivate in response to specific input values. It differs significantly from conventional processors centered around arithmetic units. The automata paradigm is capable of performing high-speed, comprehensive search, pattern-matching and analysis of complex, unstructured data streams, and shows potential to be a disruptive technology that dramatically improves processing speeds in many areas, including “Big Data,” one of the priority areas for research in U.Va.’s new strategic plan.
“We are pleased that U.Va. will be the first-in-class to lead the nation and world in demonstrating this new computing technology,” U.Va. Vice President for Research Thomas C. Skalak said. “U.Va. is committed to stand at the frontiers of knowledge and pioneer new transformational advances with benefit to society.”
About the Micron Automata Processor
Micron’s Automata Processor, announced Monday at Supercomputing 2013, an international computing conference being held this week in Denver, is an accelerator that leverages intrinsic structural parallelism of computer memory – and Micron’s expertise in manufacturing memory – to provide a massive array of automata elements, allowing massive processing parallelism with improved energy efficiency. Early applications are expected in many Big Data regimes, such as bioinformatics, video/image analytics and network security – areas that pose challenges to conventional processor architectures.
Micron Technology's revolutionary new computer chip for automata computing.Image courtesy of Micron TechnologyFor example, in one of its early projects, the Center for Automata Computing is expected to help researchers in biomedical engineering analyze huge volumes of DNA data and cellular imagery. More generally, possible applications include advancing medical research, interpreting massive economics and other social science datasets, data analysis for personal and national security, design verification in engineering and new understanding of the natural world.
With the new center, U.Va. will be working across academic departments to develop new applications for the revolutionary computer chip, as well as exploring how to improve next-generation automata design technology and programming environments.
“The automata processor has already proven far more effective in early analytics application studies compared to existing hardware, including graphic processors, and is an incredibly promising computing technology,” said Kevin Skadron, chair of the U.Va. Department of Computer Science in the School of Engineering and Applied Science, founding director of the center and one of the three lead researchers who wrote a proposal to Micron Technology for funding for the new center.
Skalak added, “The new center is very timely for a wide variety of societal challenges, from image processing to pattern recognition and cybersecurity for Internet-based activities. There is enormous potential to speed up this type of process by an order of magnitude, and that speed will have positive impact on issues in water supply, big data analytics for market projections and economic growth, health care, privacy of personal information and international trade. In turn, such advances will create new jobs.”
Automata Processor Underscores New Era of Computing
Skalak noted that the presence of a world-class technology company like Micron in Virginia, and long-standing scientific exchanges between U.Va.’s nanotechnology group – led by materials scientist Stu Wolf – and Micron’s research and development group allowed the opportunity for U.Va. to become prominent at the earliest moments of this new era in computing.
Examples of ways automata computing can be used include the ability to quickly and accurately analyze genetic data; to assess real-time imagery in such areas as medical care, pharmaceuticals, security and astronomy; and to analyze transmission patterns in epidemiology. Current computing architectures are proving incapable of performing advanced analytics on complex datasets while keeping pace with the steady stream of incoming data in a multitude of areas – the big data challenges now facing society.
“We possibly are on the verge of a computing revolution with the capabilities of automata computing, and what we learn will have a great impact on how a variety of complex problems are solved in the future,” Skadron said.
The short-term goal for the Center for Automata Computing is to quickly enhance research in applications of automata computing and to obtain early results that demonstrate the broad applicability of this paradigm, which in turn can form the basis for new collaborations.
Initial funding from Micron, together with matching support from U.Va., will cover costs for a full-time postdoctoral researcher and three graduate students. The center will serve as a hub for bringing together U.Va. computer science and engineering faculty, who can improve automata capabilities, and scientific researchers in a range of fields who can benefit from the new capabilities of automata.
Longer term, the university will bring in researchers from other Virginia universities and companies and leverage additional funding from corporations and state and federal funding agencies, and eventually will expand nationwide.
Wolf and U.Va. electrical and computer engineer Mircea Stan worked with Skadron to develop the proposal to Micron Technology for establishing the center, and will serve as associate directors.
“The center really demonstrates that U.Va. is a leader in the research of new computing technologies,” Skadron said. “Previous research we’ve done on the development of other computing technologies really laid the groundwork for our ability to work with this new paradigm.”
