Experimental Facilities Division Oak Ridge

October 13-15, 2003

Data Visualization and Analysis

Ian Anderson

Director, Experimental Facilities

October 13-15, 2003

NeSSI

Experimental Facilities Division Oak Ridge2

October 13-15, 2003

The 2003 Nobel Prize in Physiology or Medicine

Paul C Lauterbur and Peter Mansfield for their discoveries concerning

"magnetic resonance imaging“

In the beginning of the 1970s, they made seminal discoveries concerning the development of the technique to visualize different structures.These findings provided the basis for the development of magnetic resonance into a useful imaging method.

Experimental Facilities Division Oak Ridge3

October 13-15, 2003

Solvation in supercritical water

Combustion turbulence modeling

Turbulent flame

Waveguide optics

Clay-mineral geochemistry

Dissociation of ketene

Crystal structure for C36 solid

Complex fluids

Two spheres mixing in a stream

Binary alloy solidification

Office of Basic Energy Sciences

Harnessing the Power of Advanced Computing forHarnessing the Power of Advanced Computing forCondensed Matter and Materials Physics, Chemistry, and BiosciencesCondensed Matter and Materials Physics, Chemistry, and Biosciences

Electric field in a 2D photonic crystal waveguide

Cs ion transport

Gold nanowire

Nanoparticles binding in solution

Magnetic moments in materials

Na counterion mobility in DNA

Uranyl in aqueous solution

Vortices in a superfluid Semiconductor-liquid interface C-H bond activation reactionAtomic hydrogen ionization

3

                 

Experimental Facilities Division Oak Ridge4

October 13-15, 2003

Software enables science

• Instrument control• Data

formats storage accessibility ownership

• Visualization

• Treatment

• Analysis

• Complementarity

• Simulation

• Modeling

• Make neutrons a simple tool

• Cater for amateurs and pro’s

• Extract relevant information

• Optimize experiments

tanalysis < texperiment

• Present methods are outdated

• Software is often written by non- professionals

• Resources are always limited

Experimental Facilities Division Oak Ridge5

October 13-15, 2003

What has changed?

• NSSA survey (Roger Pynn)– Standardized collaborative data visualization and analysis

• DOE OS report– Deployment of software to link geographically separated

researchers, to facilitate movement of large (petabyte) data sets

• NSF S&E report– 1st priority – advanced cyberstructure – better storage, analysis,

visualization and distribution tools.

Distributed computing

• High speed computing• High speed networks• Massive data storage

Experimental Facilities Division Oak Ridge6

October 13-15, 2003

How to accomplish the Vision?

• At SNS/HFIR we have been concentrating so far on building the data acquisition software.

• The large BES software grant to Caltech (ARCS) has allowed them to start the development of an architecture (DANSE) which could be used throughout SNS and HFIR.

• Interest from other facilities and user groups to participate

• SNS must take the lead in defining the software to be used at SNS

• Have to focus the efforts at Oak Ridge!

• Short Term: Need to have software running in 3 years time

– Develop the architecture

– Incorporate legacy software

• Long term: Develop the full capability with new software

– Participation of other facilities and users

Experimental Facilities Division Oak Ridge7

October 13-15, 2003

What has been happening?

• Created a position at SNS for a task leader

• Discussions with BES in July– Very supportive of the initiative

– Want to see the impact on day 1 !!

– Asked for a detailed plan/proposal

– Asked SNS to review the DANSE proposal

• Workshop to define requirements – NeSSI (13-15 March)– Visualization tools

– Modelling

– Distributed Analysis

– Optimization of Experiments

– Data Access

– Networking

Not discussing remote control of instruments!

Experimental Facilities Division Oak Ridge8

October 13-15, 2003

Oak Ridge on the NSF Teragrid (John Cobb)