High-level software frameworks to surmount the challenge of 100x scaling for biomolecular simulation science

Next-generation exascale systems will fundamentally expand the reach of biomolecular simulations and the resulting scientific insight, enabling the simulation of larger biological systems (weak scaling), longer timescales (strong scaling), more complex molecular interactions, and robust uncertainty quantification (more accurate sampling). Since currently envisioned exascale hardware architectures are essentially larger versions of systems available today, it will be challenging to solve biological problems that require longer timescales, involve more complex interactions and robust uncertainty quantification without significant algorithmic improvements. We believe that high-level simulation algorithms incorporating high-level parallelism and leveraging the statistical nature of molecular processes can provide a means to address these challenges of scaling. Proof-of-concept simulation algorithms have yielded advanced sampling and adaptive control algorithms for efficient simulation of long timescales and complex behaviors. Novel dataflow and workflow systems are needed to implement these advanced algorithms in a way that is usable by the community in exascale systems. A middleware ecosystem that provides these in a robust, scalable, reusable, and extensible framework is a key requirement for exascale infrastructure investment to result in revolutionary biological insight.