Angik Sarkar - Information Processing with Spins and Magnets
Tuesday, November 01, 2011 12:30 PM - 1:30 PM EDT
HAAS Hall Rm. 111
Magnetic bits, which harness the spin degree of freedom of an electron, have long been used for long term and short term storage of information in hard disk and memory systems. Nevertheless, information processing has traditionally been done using charge based transistors. In this project, we consider the possibility of information processing using spins and magnets based on an all-spin logic (ASL) device. The ASL concept is based on key scientific advancements of the last decade, advancements that have blurred the distinction between spintronics and magnetics, creating the possibility of one large field with the potential of providing a low power alternative to charge-based information processing. In particular, two key recent advances are (1) the demonstration of spin injection into metals and semiconductors from magnetic contacts and (2) the switching of a second magnet by the injected spins. Magnets inject spins and spins turn magnets forming a closed "ecosystem" without the need to convert to charge. To model the 'all spin' operation, we have developed a multi-magnet simulator coupled to spin transport channels and benchmarked it against experiments. Using the model, we have shown that ASL can potentially be a low energy information processing device which can be cascaded to create complex circuits, but there are major challenges to be overcome. More interestingly, there is a possibility that this could provide a natural implementation for biomimetic systems with architectures that are radically different from the standard von - Neumann architecture that shapes today's computer design.
- Switching Energy and Delay of all spin logic devices, Applied Physics Letters, 98, 2011 All-Spin Logic Device with Intrinsic Non-Reciprocity, IEEE Transactions on Magnetic, 2011 Modeling All Spin Logic: Multi-magnet Networks Interacting via Spin Currents, IEEE IEDM, 2011