Pages

Wednesday, March 4, 2009

Summary of the Collaboratory

Summary of the Collaboratory by Ruth Pordes, Executive Director, OSG

(Notes by David Ritchie; all transcription errors are his)

Opening Slide followed by Highlights of talk
see slides for details and accurate info

OSG Consortium in existence since 2005.
OSG Project in 3rd year of 5 year funding cycle.
Core collaboration remains, strengthened by new participants and openness to new approaches.
-US ATLAS, US CMS, & LIGO with accomplishments on the ground, open collaboration to face realities and changing needs.
-Involvement of other physics increasing (CDF, D0, Alice)
-Processes of engaging new non-physics users and communities being “shaken out”.
-Leadership remains, strengthened by an excellent team and energetic contributors
Challenges and issues from detail to the national level continue - no shortage of work!

Highlights follow...

The Principles at Work

Improvments in Technologies in 2008
Integrated into Software Stack the VDT with its ~70 components

LIGO - Exemplar of Community Grid

VO throughput overpast year >40 sites contributing (see slide in talk)

20,000 CPUdays/day over the year

US ATLAS, US CMS, (US ALICE) and the WLCG

US delivered >30% of throughput robustly to the global experimentss

Science Output

CDF + D0 > 100 physics pubs in 2008

Star also

Engagement and Support Beyond Physics

Engagement team provides common point of entry for new communities
Usage naturally verye cyclic

Non-Physics VOs and Usage

Science Output

OSG asnd Satellite Partnerships --
Three: CI engagement, Structure Biology Grid, Sociology Partner

SiDAC Collaoborations / ESNET Internet 2 / DOE HPC and MPI / TeraGrid (Challenge to maintain collaboration)

DOE SciDAC/NSF Review in Jan -- Proceed

Improve and Sustain the Software Stack

Evolving to the Needs of the Future (see slides)

Somehow VO's a much more static thing -- how to really really have ad-hoc group collaborations

OSG and the National Cyberinfrastructure -- how can we help?

Questions...

Elab on small group collaborations -- SNS a science effort of interest in this area - groups form for days / months / etc. - how to deal with this and make it more possible...? Discussion

No comments: