Today at Berkeley Lab

Researchers Discuss Challenges of Storing and Retrieving Exascale Data

Future exascale supercomputers will be able to process a quintillion calculations per second. How can users store and retrieve data quickly enough on such a system? In this podcast, CRD’s Suren Byna and NERSC’s Quincey Koziol discuss prepping HDF5 – one of the most parallel I/O libraries for high-end supercomputing systems – for exascale use.