Reflections on science, technology, and computing — leavened by personal experience


It is past time to reunite the big data, cloud and high-performance computing communities. Each can each learn much from the other.

Extraordinary parallelism, unprecedented data locality and adaptive resilience: these are daunting architecture, system software and application challenges for exascale computing.

I just posted an essay on the BLOG@CACM site regarding the divergence of programming models for network (distributed) services and for parallel computing. The details are here. In the essay, MPI is a placeholder for all of the message passing libraries we have developed over the past thirty years.