Download presentation
Presentation is loading. Please wait.
Published byMavis Suzanna Holmes Modified over 9 years ago
1
PICASso PICASso Program in Integrated Computer and Application Sciences CS Finance Center Geo Astro Bio Engr. GFDL PPPL Genomics Center
2
Motivation Computational science as third pillar of science Explosion of information services Both rely on scalable parallel systems Technology marching apace, but challenge is to harness it –More complex phenomena: modeling and applied algorithms –Increasingly complex systems: computing, visualization,... Application researchers need CS understanding, collaboration –Seeking it increasingly at Princeton Many major advances in CS are increasingly application-driven –CS must deal with scale and complexity of real applications
3
Motivation (contd.) Next-generation researchers need to cross discipline boundaries –progress in own primary research areas applications, algorithms, systems, visualization –new research challenges driven by new applications or technologies –new research areas developed at boundaries of existing ones Lack of “bridge” people is hindering progress –especially in computational sciences –need to train such bridge people
4
Overall Goal Integrated Research and Training in entire computational pipeline Problems/Models Algorithms Parallel Software Parallel Systems Data Visualization & Management Analysis Tools
5
Specific Goals Research: Foster interdisciplinary research on campus –Between other departments and CS –Between other departments themselves Training: –Train science and engineering students in relevant aspects of CS –Train new generation of application-driven computer scientists –Train new breed of researcher: creates boundary areas, builds bridges Related Impact: –Add value to, and attract best students to, CS and other departments –Develop a successful program that others will emulate –Nucleus of a Computer and Application Sciences Center (?) –Hasten the simulation revolution in science
6
Why at Princeton? Why Now? Excellent science/application departments Growth in applications of scalable computing Existing cross-disciplinary collaborations Strong demand from other departments, undergrads, postdocs Synergy with new Centers (Genomics, Finance) and local labs Synergy with existing programs: PACM, PAC (undergrad) Rapid change, and clear recognition of need Keep computational science exciting for CS students
7
Existing Interdisciplinary Collaborations Computational Biology –Singh (CS) and Weigert (MolBio): simulating immune response –Singh (CS) and Altman (Stanford Bio): protein structure Biological Computing –Lipton (CS) and Landweber (MolBio) Computational Cosmology –Singh (CS) and Ostriker (Astro) Visualization –Dobkin (CS), Finkelstein (CS), Ostriker (Astro), Cen (Astro) –Display Wall (CS) with PPPL, Astro Labs: GFDL, PPPL, AT&T, Lucent...
8
Support and Participation NSF IGERT grant (in approval process) Open to students/advisors from all departments –in proposal: CS, Astro, MolBio, EEB, Geo, Plasma Physics –plus others: Engineering departments, Genomics, Finance, etc. CS Finance Center Geo Astro Bio Engr. GFDL PPPL Genomics Center “Hub-and-spoke” model centered in Computer Science
9
Focus Areas Computer Science –Models and methods –Parallel and distributed computing –Visualization and data management Applications –Keep range of application areas wide and flexible –Natural sciences, engineering, finance and commercial, … –Scalable information services Not traditional numerical analysis or Scientific Computing program Research at boundary of applications and Computer Science (algorithms and systems)
10
Cross-Cutting CS Research Thrusts New models –esp. for complex dynamic systems; e.g. immunology Algorithms (sequential and parallel) –e.g.space- and time- adaptive methods Tightly-coupled scalable parallel computing –algorithms, scaling, programming models, performance portability, programming environments, performance prediction Parallel computing on clusters and “systems of systems” –room-wide and campus-wide Visualization –extracting and steering through meaningful data –immersive visualization using display wall All areas highly application-driven
11
Initial Application Research Thrusts Astrophysics –Galactic structure formation and nature of matter –Analyzing observational data: Sloan Sky Survey Biology –Protein structure determination –Simulation of complex dynamic systems –Immunology, Neurobiology, Genomics Geosciences –Earth’s surface circulation systems (ocean, atmosphere) –Earth’s interior circulation (solid-state deformation) Information Management –scalable networked services
12
Structure and Infrastructure Executive Committee –Jeremiah Ostriker (University Provost and Astro) –Jaswinder Pal Singh (CS) –Kai Li (CS) –David Dobkin (CS Chair) Others in Managing Committee (one per department) –Weigert (MolBio), Spergel (Astro), Bunge (Geo), Held (GFDL), Tang (PPPL), Levin (EEB), Research Scientist Equipment –64-proc. Origin2000 (Astro/CS/PPPL), more to come –128-proc. production cluster, 64-proc. development cluster –Display Wall –High-end rendering equipment (CS/Astro)
13
Key Features Interdisciplinary curriculum development –CS courses in Computer Systems and Visualization –New computational courses in other departments –Team-taught cross-disciplinary vertical project courses –Tutorials and Seminars Cross-department joint advising Integrated research groups across disciplines Internships in other departments and laboratories/industry Cross-cutting annual thematic programs Mechanisms to bring people together –“PICASso Central” in CS building –Computational Research in Princeton seminar series
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.