Abstract:
For science and technology to continue progressing at the current pace, we need our computational processing capabilities to keep growing. In conventional transistor-based computers, high heat dissipation and slow transfer rates between processors and memory are bringing about the breakdown of Moore’s Law, which forecasts an exponential increase in computational speed. Computing with molecules is an attractive option given the vast chemical space and relatively low energy dissipated in chemical reactions compared to transistors. In this talk, I will present a new approach for molecular computing using a network of chemical reactions taking place within an array of spatially localized microdroplets. Each droplet represents one bit of information and combinatorial optimization problems are encoded in the form of intra- and inter-droplet interactions. The problem is solved by initiating the chemical reactions within the droplets and allowing the system to reach a steady-state; in effect, we are annealing the effective spin system to its ground state. I will explain the theory behind this approach and show how this molecular computer can be used to solve hard problems such as Boolean satisfiability, traveling salesperson, and lattice protein folding. I will also discuss the requirements for its physical implementation and outline a roadmap to achieving a purely molecular computer that relies on pre-programmed nearest-neighbour inter-droplet communication via energy or mass transfer. Finally, I will show some recent experimental results of a hybrid classical-molecular computer, a stepping-stone device that allows us to test the chemical reactions underlying the computation.