Extreme Scalability Working Group

From TeraGrid Wiki

1 Motivation and Purpose
2 Extreme Scalability Workshop at TeraGrid '08
3 Fault Tolerance for Extreme-Scale Computing
4 Meetings
- 4.1 Teleconferences
- 4.2 Extreme Scalability Working Group meetings at TG08
5 Documents and Links
6 Resource Guide for Extreme Scale Computing
7 Mailing List
8 PetaScale Funded Projects

Motivation and Purpose

The purpose of the Extreme Scalability Working Group (XS-WG) is to meet the challenges and opportunities of deploying extreme-scale resources into the TeraGrid, maximizing both scientific output and user productivity.

Specifically, the XS Working Group addresses the following technical challenge areas:

Scalability and architecture (algorithms, numerics, multicore, etc.)

POC: Raghu Reddy (PSC)
Members: Reddy, Majumdar, Urbanic, Kim, Lynch, Jha, Nystrom

Tools (performance, debugging, compilers, etc.)

POC: Mahin Mahmoodi (PSC)
Members: Mahmoodi, Wright, Alameda, Shende, Sheppard, Brown, Nystrom

Workflow, data transport, analysis, visualization, and storage

POC: Phil Andrews (NICS)
Members: Andrews, Welling, Nystrom, end-to-end person TBD, workflow person TBD

http://www.teragridforum.org/mediawiki/index.php?title=Visualization#Petascale_Visualization_and_Data_Analysis_White_Paper

Action items for the TeraGrid

POC: Nick Nystrom and Sergiu Sanielevici (PSC)
Members: Nystrom, Sanielevici, Lathrop, Dudek, Bolte, user-facing projects person TBA, allocations person TBA

If you would like to participate in the XS Working Group, contact Nick Nystrom, XS-WG Lead.

[edit]

Extreme Scalability Workshop at TeraGrid '08

A Extreme Scalability Workshop will be held on Monday, June 9, in Las Vegas, coinciding with the start of TeraGrid '08. This workshop is a follow-on to NSF's Building Petascale Applications and Tools on the TeraGrid workshop, which was held in Tempe on December 11-12, 2007.

Our current plan is to focus in detail on two topics, each for one half day. In the first half, people representing applications and tools will present requirements, plans, and open questions, followed by ample time for discussion of how requirements may be met, where additional work is needed, and which groups need tighter interaction. In the second half, we will dive into multicore and other architectural issues and their relationships to applications and tools, again with ample time for discussion.

Registration for this workshop will open soon; please check back for instructions. Participants are also encouraged to attend TeraGrid '08, which will feature a related BoF, Petascale science and user support, as well as technical papers, posters, and valuable networking.

If you have any questions about this workshop, contact Nick Nystrom.

Slides and notes from the workshop.

[edit]

Fault Tolerance for Extreme-Scale Computing

A follow-up workshop, focusing on fault-tolerance, was held in March 2009.

The purpose of this workshop was to discuss fault-tolerance on large systems for running large, possibly long-running applications. The main point of the workshop wa to have systems people, middleware people (include FT experts), and apps people talk about the issues and figure out what needs to be done, mostly at the middleware and app levels, to run such apps on the coming petascale systems, without having faults cause large numbers of application failures.

2009 Fault Tolerance Workshop

[edit]

Meetings

[edit]

Teleconferences

The XS Working Group currently meets by teleconference on Wednesdays at 3PM Central time. Once everyone is fully engaged and making solid progress, we expect to back off to biweekly telecons.

Minutes: [ 3/5/08 | 3/12/08 | 3/19/08 | 3/26/08 | 4/2/08 | 4/16/08 | 4/30/08 | 5/14/08 | 5/28/08 | 6/12/08 | 7/9/08 | 8/20/08]

[edit]