Sharad Singhal, The Machine & Matthias Becker, University of Bonn | HPE Discover Madrid 2017
summary
Description of a working memory‑driven system that places data at the center with 160 terabytes of globally addressable memory across about 40 compute nodes, used to analyze a 30,000‑person, 30‑year Alzheimer’s cohort (genomes, blood, high‑res MRIs). An emulated stack on Superdome X and code changes cut a genomics pipeline from six days to 22 minutes then to 13 seconds, showing concrete speedups, tools, and real‑world end‑to‑end results.
AI-written summary · about
