Exploring the readability of nano-magnetic energy minimizing co-processor

Recent times have witnessed profound interest to find a suitable hardware platform for non-boolean computation. Our approach relies on the fact that the Hamiltonian of a system of coupled nanomagnets is quadratic. A wide class of quadratic energy minimization can be solved directly by the relaxation of a grid of nanomagnets [1]. Preliminary research showed that magnet-based solutions are independent of problem size. The magnetic states of the nano-magnets provide solution of the original optimization problem. Since the framework is similar to STT-MRAM memories, we can leverage the heterogeneous integration with CMOS technologies for access, control and reading the cells. This paper focuses on an efficient reading of the magnetic cells which is of critical importance. Based on the framework, the reading mechanism has to be able to differentiate the two states (inplane single domain and vortex domain states). We investigated two reading mechanisms for the detection of relaxed magnetic states. Our comparative study shows that differential read scheme is more capable to handle the effects due to the process variation.