Milo Polte
CMU Student

Garth Gibson
CMU Instructor

John Bent
Mentor

• ...And eat it too Paper
• Supercomputing 2009

This project explores how well suited data intensive computing programming/run time paradigms like map reduce and other graphs apply to scientific applications.

Julio Lopez
CMU Instructor

Gary Grider
Mentor

James Nunez
Mentor

John Bent
Mentor

Salman Habib
Mentor

Thomas Brettin
Mentor

Pavel Senin
Mentor

Tony Heaton
Mentor

Michael Fisk
Mentor

Laura Monroe
Mentor

• Introducing Map-Reduce to High End Computing PDSW08 talk
• paper
• PDL 2009 PLFS and HDFS Poster

The concept is to head HPC file system storage towards file formats in the file system.  It is quite possible that many file types like N processes to 1 file with small strided writes might be served well by special handling at the file system level.  Decades ago, file types and access methods were used and were supported within a single file system.  The IBM MVS storage systems allowed for many different file types, partitioned data sets, indexed sequential, virtual sequential, and sequential to name a few.  Storage for modern HPC systems may benefit from a new parallel/scalable version of file types.  There is much research to be done in this area to determine the usefulness of this concept and how such a thing would work with modern supercomputers and future HPC languages and operating environments.

Milo Polte
CMU Student

Garth Gibson
CMU Instructor

Gary Grider
Mentor

James Nunez
Mentor

John Bent
Mentor

• The related LANL Parallel Log Structured File Systems PLFS
• Efficient Data Placement in a Hybrid Storage Architecture
• Large-scale Evaluation of GIGA+ Scalable Directories
• Comparing Performance of Solid State Devices and Mechanical Disks PDSW08 talk
• Log-structured Files for Fast Checkpointing
• Network File System (NFS) in High Performance Networks
• usenix paper
• SOSP paper
• Sigmetrics paper
• Full Paper
• IDCE paper
• TOS paper
• paper
• PDL Day PLFS Poster
• PDL Day SSD Poster
• PDL Day Talk
• PDL Day PLFS talk
• PLFS Poster
• PDL 2009 Poster
• PDL 2009 Presentation

Computer scientists at LANL have become intimately involved in studying the software development challenges posed by the new Roadrunner cell hybrid architecture, as well as possible approaches to addressing these. One of their findings is that the core structure of existing applications has to be revisited and understood before they can be effectively transitioned to a new architecture. Understanding this structure requires analysis tools to assess time, IO, memory usage, and other operational characteristics of the application so that its data, functions and memory patterns can be mapped to appropriate segments of the new processor architecture.

The purpose of this project is to study the OpenSpeedShop tool framework and identify a set of constraints it imposes on tool plugins, drawing on experience from OpenSpeedShop developers at LANL.  The goal is to understand the rationale behind these constraints in terms of framework quality attributes like performance and extensibility, and to explore potential design alternatives and their tradeoffs with respect to those quality attributes.

Ciera Jaspan
CMU Student

Jonathan Aldrich
CMU Instructor

Dave Montoya
Mentor

Steve Painter
Mentor

• ASE Paper 2008
• ECOOP Paper 2009
• RAOOL Paper 2009
• Checking Framework Interactions with Relationships

Garth Gibson
CMU Instructor

Greg Ganger
CMU Instructor

David Andersen
CMU Instructor

HB Chen
Mentor

Andrew Shewmaker
Mentor

Parks Fields
Mentor

• FAST09 WIP Abstract