Scaling I/O Beyond 100,000 Cores using ADIOS
14 Jan 2014 - 14 Jan 2014

Date: 14 January 2014, Tuesday
Time: 9 am – 12pm
Venue: Level 17, Charles Babbage Room, 1 Fusionopolis Way,Connexis South, Singapore 138632

Speaker: Scott A. Klasky, Oak Ridge National Laboratory


Supercomputers are capable of performing 10^16 floating-point operations per second (34 PFlops). The greatest challenges facing computer and computational scientists are to further increase the computer speeds and, more challengingly, to develop programming models to enable realization of the potentials of such massive systems.

We will discuss three related topics. First, we review the latest innovations in interconnection networks and processor technologies for achieving ever increasing raw computing speeds. Second, we analyse a parallel computing algorithm, task mapping, for helping minimize and balance data movement over a complex network of processors. Third, we discuss several computational science projects including study of the mechanisms of human blood platelet activations, causing heart attacks and strokes, by multi-scale discrete particle dynamics and molecular dynamics.

As concurrency continues to increase on high-end machines, from both the number of cores and storage devices, we must look for a revolutionary way to treat Input/Output (I/O).  As a matter of fact, one of the major roadblocks to exascale is how to write and read big datasets quickly and efficiently on high-end machines.  On the other hand, applications often want to process data in an efficient and flexible manner, in terms of data formats and operations performed (e.g., files, data streams).  We will show how users can do that and get high performance with ADIOS on 100,000+ cores.  Part I will introduce parallel I/O and the ADIOS framework to the audience. Specifically, we will discuss the concept of ADIOS I/O abstraction, the binary-packed file format, and I/O methods along with the benefits to applications. Since 1.4.1, ADIOS can operate on both files and data streams. Part II will include a session on how to write/read data, and how to use different I/O componentizations inside of ADIOS. Part III will show users how to take advantage of the ADIOS framework to do compression/indexing. Finally, we will discuss how to run in-situ visualization using VisIt/Paraview+ADIOS. 



Scott A. Klasky is the group leader for Scientific Data in the Computer Science and Mathematics Division at the Oak Ridge National Laboratory. He holds a Ph.D. in Physics from the University of Texas at Austin (1994), and has previously worked at the University of Texas at Austin, Syracuse University, and the Princeton Plasma Physics Laboratory. Dr. Klasky is a co-author on over 180 papers, and is the team leader of the Adaptable I/O System (ADIOS), which won an R&D 100 Award in 2013.

No of Participants:  38