Exploring models of computation and the consequences for systems for Big Data

Peter Hofstee

IBM Austin Research Laboratory

This talk is an attempt to come to grips with fundamental distinctions underlying the different computational frameworks for Big Data. Our main goal is to better understand how to optimize Big Data systems. We also believe our approach may aid in teaching. We show how common approaches like Hadoop, SPARK, MPI, and OpenMP and Streams relate. Finally we discuss a system built on OpenPOWER technology that leverages flash as memory. We show how such a system might fit our framework and we discuss the system-level benefits that such an approach can bring.



Biography: Dr. Peter Hofstee currently works at the IBM Austin Research Laboratory on workload-optimized and hybrid systems. Peter has degrees in theoretical physics (MS, Rijks Universiteit Groningen, Netherlands) and computer science (PhD, California Inst. of Technology). At IBM Peter has worked on microprocessors, including the first CMOS processor to demonstrate GHz operation (1997), and he was the chief architect of the synergistic processor elements in the Cell Broadband Engine, known from its use in the Sony Playstation 3 and the Roadrunner supercomputer that first broke the 1 Petaflop Linpack benchmark. His interests include VLSI, multicore and heterogeneous microprocessor architecture, security, system design and programming. Peter has over 100 patents issued or pending