The storage, network and compute needs of these projects are really exciting. You get a couple gigs to a couple dozen gigs per measurement run, and you're running a lot of measurements. As the presentation shows, they are looking at a petabyte per month. That's a lot of stuff to crunch. CERN is also pushing 2 - 3 digit petabyte storages on ceph around for their LHC.
Genome analysis has similar interesting requirements. A sequencer for a human genome outputs something like 100 - 200Gb of raw data to store and process, and then you have some computationally tricky analysis to run on that.
Genome analysis has similar interesting requirements. A sequencer for a human genome outputs something like 100 - 200Gb of raw data to store and process, and then you have some computationally tricky analysis to run on that.