Peter M. Kogge, McCourtney Prof. of CSE, Univ. of Notre Dame
Thursday, May 23, 4pm Annenberg 105
“Big Data” traditionally refers to some combination of high volume of data, high velocity of change, and/or wide variety and complexity of the underlying data. Solving such problems has evolved into using paradigms like MapReduce on large clusters of compute nodes. More recently, a growing number of “Deep Data” problems have arisen where it is the relationships between objects, and not necessarily the collections of objects, that are important, and for which the traditional implementation techniques are unsatisfactory.
This talk addresses a study of a class of such “challenge problems” first formulated by David Bayliss of LexisNexis, and what are their execution characteristics on both current and future architectures. The goal is to discover, to at least a first order approximation, what are the tall poles preventing a speedup of their solution. A variety or architectures are considered, ranging from standard server blades in large scale configurations, to emerging variations that leverage simpler and more energy efficient chip sets, through systems built on 3D chip stacks, and on to new architectures that were designed from the ground up to “follow the links.” Such architectures are considered for two variants of such problems: a traditional partitioned data approach where data is “pre-boiled” to provide fast response, and one that uses very large graphs in very large shared memories.
The results are not necessarily intuitive; the bottlenecks in such problems are not where current systems have the bulk of their capabilities or costs, nor where obvious near term upgrades will have major effects. Instead, it appears that only highly scalable memory-intensive architectures offer the potential for truly major gains in application performance.
Image courtesy Amber Harmon.
Computing pioneer, renowned computer scientist, and former CACR Director Paul Messina has been presented with the distinguished Thomas Hart Benton Mural Medallion at Indiana University. “Paul Messina, Indiana University salutes you. Throughout your distinguished career, your work helped lay the foundation for grid and cloud computing, and helped bring parallel computing technology to the forefront of scientific computing,” said IU President Michael McRobbie.
“Paul established the effective use of high-performance computing at Caltech, and was instrumental in showing the broader research community its potential for advancing science at the national scale. His legacy lives on at CACR to this day, for which we are grateful, and we salute his accomplishments” said Mark Stalzer, CACR’s current director.
(Read more at http://www.isgtw.org/spotlight/computing-pioneer-honored-big-red-ii-dedication)
Thursday April 11, 2013
Powell-Booth Room 100
“Turning Large Simulations into Numerical Laboratories”
Alex Szalay, Alumni Centennial Professor, Department of Physics and Astronomy, The Johns Hopkins University
The talk will discuss how large (100TB+) supercomputer-scale simulations can be turned into interactive public laboratories. Examples include simulations of turbulence and various cosmological simulations, soon to reach the PB scale.
The central disk of our Milky Way galaxy with the Gemini stellar stream highlighted in the top right of the image. Credit: Andrew Drake/Axel Mellinger
A stream of at least 150 ancient variable stars has been confirmed to extend some 130,000 light years beyond our own galaxy’s stellar halo — on the fringes of the Intergalactic Medium, where aside from hot gas and dark matter, space-time becomes as sparse as the deep Sahara. The confirmation, based on analysis of 10 billion year-old dying RR Lyrae stars in the Gemini constellation, was done by an international team of astronomers, including CACR scientist Andrew Drake, who report their findings in The Astrophysical Journal.
(Read the full story)
Jan 31, 2013 1:00 PM
Keith Spalding 410
Andrew Drake, CACR
We have performed an extensive search for RR Lyrae among the 500 million sources observed by the Catalina Surveys. We detect ~26,000 type-AB RR Lyrae (of which 20,000 are new discoveries) from a region spanning 3/4 of the sky. By determining accurate distances to the stars, we investigate the spatial distribution of structures within the Milky Way halo. Combining the RR Lyrae distances with SDSS spectroscopy we are able accurately trace the velocities and metallicities of hundreds of sources within the Sagittarius tidal streams system. We find the first strong evidence for a dense tidal stream that overlaps the Sagittarius system to distances beyond 100kpc, yet remains unexplained by any existing model.