Programming a Parallel Future

Things change fast in computer science, but odds are good that they will change especially fast in the next few years. Much of this change centers on the shift toward parallel computing. In the short term, parallelism will thrive in settings with massive datasets and analytics. Longer term, the shift to parallelism will impact all software. In this note, I’ll outline some key changes that have recently taken place in the computer industry, show why existing software systems are ill-equipped to handle this new reality, and point toward some bright spots on the horizon.