Events and Meetings
Bay Area Beowulf User Group (BayBUG)
April 17, 2007
2:30 - 5:00 p.m.
AMD headquarters Common Building,
991 Stewart Drive, Sunnyvale
Join moderator and Beowulf cluster co-inventor Donald Becker for food and drinks and to learn from and network with other Linux HPC professionals.
- Chris Smith, Platform Computing
- Dr. Greg Benson, University of San Francisco
Title: Job Management Standards in HPC
By Christopher Smith
Standards are often touted as instant cure to system interoperability issues, but what is the reality compared with the hype? This talk will provide an overview of current standards activity at the Open Grid Forum pertaining to job management within HPC schedulers. Benefits of standards for schedulers will be discussed, API versus protocol level standards will be contrasted, and the results of some recent interoperability demonstrations at SC06 will be reviewed.
Chris Smith is a Senior Product Architect at Platform Computing. Over his last 9 years at Platform, he has focused on the integration of Platform's Grid middleware into production Grid solutions within High Performance Technical Computing disciplines. He is also an active participant in the Open Grid Forum, being a contributor on the JSDL specification, and a co-author on both the OGSA BES and HPC Profile specifications.
Title: Rapid Development of Parallel Systems and Applications in River
By Dr. Greg Benson
River is a parallel and distributed programming environment written in Python. The River core interface is based on a few fundamental concepts that enables the execution of code on multiple virtual machines and provides a flexible mechanism for communication. These concepts are supported by the River run-time system, which manages automatic discovery, connection management, and naming. River can be used directly by an application programmer to implement parallel programs or it can be used as a framework for implementing programming models. We have found the simplicity and elegance of the River core combined with Python's dynamic typing and concise notation make it easy to rapidly develop a variety of parallel applications and run-time systems. In this talk I will give an overview of the River system and the core River interface. I will also summarize some of the programming models we have implemented in River including a new, simple interface for task farming called Trickle and an implementation of MPI.
Greg Benson is an associate professor and chair in the Department of Computer Science at the University of San Francisco. Greg's research areas include operating systems, parallel computing, and programming languages. He has designed and developed several run-time systems and tools for parallel programming languages and libraries. He led the development of USFMPI, a multi-threaded implementation of MPI 1.2 for Linux using either Myrinet or Ethernet. Greg is a co-creator of FlashMob Computing and he implemented much of the software that enabled the harnessing of 700 volunteer computers in a single day at the USF gym to run HPL (Linpack). Greg received his Ph.D. in computer science from the University of California, Davis. While he was a graduate student, he held visiting research positions with Orca research group at Vrije University in the Netherlands and with the Flux OS research group at the University of Utah.
Nov. 14, 2006 BayBUG Podcast
"Commercialization of Open Source - for better or worse?"Moderator:
- Doug Eadline, Editor in Chief of ClusterMonkey
- Donald Becker, CTO of Penguin Computing, Beowulf cluster co-inventor
- Thomas Sterling, Faculty Associate, Center for Advanced Computing Research, California Institute of Technology (Caltech) Beowulf cluster co-inventor
- Mike Fitzmaurice, HPC Technologist at GTSI Corp.