Micron develops new parallel processing architecture

Micron has announced the development of a computing architecture capable of performing high-speed, comprehensive search and analysis of complex, unstructured data streams.

The company’s Automata Processor (AP) is an accelerator that leverages the intrinsic parallelism of memory and aims to advance computing capabilities in areas such as bioinformatics, video/image analytics and network security, which pose challenges for conventional processor architectures because of the amount of complex, unstructured data.

Automata processing exploits the natural parallelism found in Micron’s semiconductor devices to deliver a level of parallelism previously unattainable with legacy architectures.

Unlike conventional CPUs, the AP is a computing fabric comprised of tens of thousands to millions of processing elements interconnected to create a task-specific processing engine capable of solving problems with unprecedented performance.

“The Automata Processor is a breakthrough technology that is designed to use advanced memory-based processing to solve complex computing challenges that existing solutions are not able to tackle effectively,” said Chirag Dekate, IDC research manager, HPC. “This technology has the potential to solve some of the world’s most complicated data-intensive challenges, including real-time security that could dramatically affect antiterrorism efforts, or the highly efficient analysis of complex plant genomes that could allow scientists to rapidly advance their research agendas beyond what is possible today.”

Micron is working closely with ecosystem partners and research institutions to grow awareness and engagement for this new technology. “By deploying this in interesting ways, we have been able to solve a much larger instance of the NP-hard biological motif-finding problem than was previously reported, using the resources within a single Automata Processor board,” said Srinivas Aluru, professor of computational science and engineering at Georgia Institute of Technology and a leader in the field of high-performance, computational biology.

Michela Becchi, an assistant professor in the Department of Electrical and Computer Engineering with joint appointments in Computer Science and the Informatics Institute of the University of Missouri, has focused her efforts on using the AP to address the challenges involved with implementing high-speed regular expression matching engines. The processor, said Becchi, efficiently supports large sets of complex regular expressions, while providing worst-case processing guarantees and support for multiple input streams. In addition, it comes with a user-friendly programming tool-chain that can be easily integrated with existing tools for regular expression processing.