ALBANY, N.Y. — A SUNY Polytechnic Institute (SUNY Poly) professor will use a grant of $1.2 million for a research and development project on next-generation computer systems.
Nathaniel Cady, professor of nanobioscience, will use the Air Force Research Lab (AFRL) funding. It is part of a $2.4 million grant in collaboration with the University of Tennessee, Knoxville.
The funding seeks to enable the fabrication and testing of an adaptive neural network array based on memristors.
SUNY Poly describes memristors as “technology which could one day lead to faster, more energy efficient and powerful computer capabilities.”
The three-year AFRL award enables the development of a memristor-based dynamic adaptive neural network array (mrDANNA).
As part of the collaboration, researchers will design the mrDANNA at Tennessee and will leverage SUNY Poly’s Albany NanoTech Complex to create chip prototypes.
The project will test and refine the chip prototypes to address “contemporary application challenges” and enable continued computer-chip performance scaling, especially as researchers expect Moore’s Law to approach “physical limitations.”
Moore’s Law says that “processor speeds, or overall processing power for computers will double every two years,” according to mooreslaw.org, a website that seeks to explain the law.
As the acronym indicates, the mrDANNA research is based on memristors, technology that can take the 1’s and 0’s of transistor-based computing to the next level by utilizing more than the binary on/off states that are used in today’s computers.
Memristors are able to change their resistance values incrementally; the greater the number of resistance levels, the more bits of information that each memristor can store.
Placing them in an adaptable, “neuromorphic” array opens the door to computers that would more closely resemble neurons and synapses, or the “complex” hardware of the human brain.
As part of this new effort, the team of SUNY Poly researchers plans to optimize memristor electrical characteristics to “achieve peak performance” of the neuromorphic array so that it offers the best real-time “dynamic adaptability, configurability, and scalability.”
In addition, with tightly packed memristors, they offer “greater” device density, which enables “increased” energy efficiency as well, according to the news release.
Reaction
“The Air Force Research Lab grant underscores the fact that Gov. Cuomo’s strategy catalyzing a high-tech research and development ecosystem in New York state, pioneered by SUNY Poly, is driving development of technologies which have the potential to revolutionize computing in the years ahead,” Michael Liehr, executive VP of innovation and technology and VP of research at SUNY Poly, contended.
“We are grateful to the AFRL for recognizing this trailblazing work, and we look forward to collaborating with the research team at the University of Tennessee to substantially improve computer memory. Another key benefit is that SUNY Poly students will observe and be a part of this research, providing a truly one-of-a-kind educational opportunity,” Cady said in the release.
Contact Reinhardt at ereinhardt@cnybj.com
