Scientific Computing Research Focus Group An Initiative of the Bagley College of Engineering

1
Scientific ComputingResearch Focus Group -An
Initiative of the Bagley College of Engineering -

• Ioana Banicescu
• Department of Computer Science and Engineering
  and
• Center for Computational Sciences
• Mississippi State University

2
Overview

• Mission, Initiative, Goals
• Purpose of research focus group
• Benefits and contributions
• Strategy -
• Case study - Prof. Seong-Gon Kim
• Conclusions

3
Mission

• Promote scholarship through multidisciplinary
  research in CompSci and CompEng
• Increase visibility MSU in Scientific Computing
  on national and international arenas

4
Initiative and Goals

• What entity would we like to be?
• Interdisciplinary
• Address computational challenges of complex
  problems in science and engineering through
  state-of-the-art models, algorithms, techniques
  and tools
• Initiative already exist through CCS and other
  research centers on campus SimCenter, CAVS, PET,
  GRI, Viz, LSBI, NSTI, etc.
• Find common denominator for scientific models,
  algorithms and methods, that could be used across
  research areas and disciplines.

5
Potential Participants
CCS
SimCenter
CAVS
GRI
DoD PET
Viz
LSBI
BCoE AS
NIH BioMed
Other Dept..
6
Initiative and Goals (cont)

• Why is it important?
• Provide a nucleus for "new frontier" knowledge
  and research and an awareness of existing
  expertise and capabilities
• Provide an mechanism to effectively asses and
  disseminate this knowledge, to enhance the
  intellectual resources and promote research
  collaboration across campus.
• Act as an incubator that would provide a fast
  selective process in nurturing and promoting new
  knowledge and research ideas.

7
Purpose of research focus group

• Define current research strengths in areas of
  significance to CompSci and CompEng or to other
  scientific areas with direct impact on them.
• Identify faculty members (established strength,
  recognition and support) that could act as
  attractors to enhance scientific computing at
  MSU.

8
Purpose of research focus group (cont)

• Identify research areas, approaches and questions
  on which MSU might strategically build for
  recognition in scientific computing
• Identify barriers in scientific computing
  expertise and capability that restrict us from
  developing our full potential in CompSci and
  CompEng .
• Identify funding opportunities for research that
  would enable pursuing the research goals
  mentioned.

9
Purpose of research focus group (cont)

• Identify strategies, faculty and resources that
  could enhance recognition of our scientific
  computing research and increase its
  competitiveness.

10
Benefits

• Need to effectively accomplish the following
• enhance individual's research program (substance
  and funding) (I would like to get more
  knowledge/publications/funding than otherwise I
  would).
• enhance individual's collaborations, potential
  for collaboration
• enhance overall individual's research recognition
  (in the specific area of expertise, as well as in
  the area of "impact").

11
Contributions

• Talks, meetings, (provide ideas)
• Comment on "matches" and implications
• Participate in formulating conclusions keeping in
  mind our mission and goals

12
Case Study

• Prof. Seong-Gon Kim
• "The meaning of a multidisciplinary collaborative
  effort"

13
Conclusions

• Comment on your research interest, needs,
  your expectation from participation, and your
  possible contributions to this focus research
  group.
• Administrative matters
• distinguished colloquium speaker
• meeting times, frequency
• length of talks
•  The ERC - core logistics and capabilities to
  make it happen

14
ERC HPC resources

• The ERC provides substantial high performance
  computing resources for use by its member
  centers, including
• 384 processor Intel Xeon cluster with InfiniBand
  interconnect,
• 1038 processor Intel Pentium III cluster,
• 64 processor SGI ORIGIN 2000,
• 64 processor Sun UltraSPARC cluster, and
• numerous additional special purpose machines.

15
ERC HPC resources (cont)

• The ERCs high-end visualization needs are met by
  an
• immersive CAVE-like virtual reality environment,
• two SGI Onyx2 graphics servers, and a 20 node
  visualization cluster.
• An access grid node allows for state-of-the-art
  video teleconferencing and remote collaboration.