In September of 2007, a privately held Boston company, Allied Minds, became enthusiastic about some NYU physics research when they visited the NYU Office of Industrial Liaison. Its object of interest: research in magnetism by Professor Andrew Kent, an award-winning scientist and a fellow of the American Physical Society who is an inventor on several patents based on his work at NYU. The technology which caught the investors’ eye, though it was still in development, was a spin transfer technology that could be used for high performance computer memory, involving a way to switch the direction of magnetic vectors (when the vectors point one way they represent a one; the other way, a zero). Before Kent’s discovery, magnetic memory devices used parallel or collinear orientation of the magnetization vectors. Professor Kent’s discovery was that a perpendicular orientation offers much better, faster switching performance—while using far less energy. "We were quite attracted to the opportunity. Computer memory is a very large market," says Vincent Chun, vice president of Allied Minds and general manager of Spin Transfer Technologies.
Allied Minds made an investment in December of 2007, establishing Spin Transfer Technologies and entered into a licensing deal and a sponsored research agreement with NYU. This provided equipment, enlarged Kent’s lab space, allowed for fabricating new test devices, and provided support for three post-docs. "We hope to transform the post docs now doing research into employees one day," says Chun. For now, it's still a startup company in the early stages.
Recent highlights: Spin Transfer partnered with a German manufacturer that deposits magnetic materials onto silicon wafers. It's also working with a foundry service. "These collaborations have led to excellent results in terms of high-speed switching and energy required," Chun says. The technology has also met or exceeded the majority of key performance metrics for its class, set by DARPA (the Defense Advanced Research Projects Agency). "Additional development work is needed," says Chun, but in 2010, "we’re much closer to commercialization."
