Science is ‘going virtual’ in a big way: rapidly escalating computing horsepower is turning lots of real-world research into simulation-based experimentation. Is this helping or hindering the latest open science initiatives?