changing science and engineering: the impact of hpc sept 23, 2009 edward seidel assistant director,...

7
Changing Science and Engineering: the impact of HPC Sept 23, 2009 Edward Seidel Assistant Director, Mathematical and Physical Sciences, NSF (Director, OCI)

Upload: laurel-linda-palmer

Post on 31-Dec-2015

215 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Changing Science and Engineering: the impact of HPC Sept 23, 2009 Edward Seidel Assistant Director, Mathematical and Physical Sciences, NSF (Director,

Changing Science and Engineering: the impact of HPC

Sept 23, 2009Edward SeidelAssistant Director, Mathematical and Physical Sciences, NSF (Director, OCI)

Page 2: Changing Science and Engineering: the impact of HPC Sept 23, 2009 Edward Seidel Assistant Director, Mathematical and Physical Sciences, NSF (Director,

Profound Transformation of ScienceGravitational Physics

Galileo, Newton usher in birth of modern science: c. 1600

Problem: single “particle” (apple, planet) in gravitational field (General 2 body-problem already too hard)

MethodsData: notebooks (Kbytes)Theory: driven by dataComputation: calculus by hand (1 Flop/s)

Collaboration1 brilliant scientist, 1-2 student

Page 3: Changing Science and Engineering: the impact of HPC Sept 23, 2009 Edward Seidel Assistant Director, Mathematical and Physical Sciences, NSF (Director,

3

• 3D Collision

• Science Result

• Year: 1998

• Team size

• ~ 15

• Data produced

• ~ 50Gbytes

• 3D Collision

• Science Result

• Year: 1998

• Team size

• ~ 15

• Data produced

• ~ 50Gbytes

Profound Transformation of ScienceCollision of Two Black Holes

• Science Result

• The “Pair of Pants”

• Year: 1994

• Team size

• ~ 10

• Data produced

• ~ 50Mbytes

• Impact of HPC taking root

• Science Result

• The “Pair of Pants”

• Year: 1994

• Team size

• ~ 10

• Data produced

• ~ 50Mbytes

• Impact of HPC taking root

Science ResultThe “Pair of Pants”

Year: 1972 Team size

1 person (S. Hawking) Computation

Flop/s Data produced

~ Kbytes (text, hand-drawn sketch)

400 years later…same!

Page 4: Changing Science and Engineering: the impact of HPC Sept 23, 2009 Edward Seidel Assistant Director, Mathematical and Physical Sciences, NSF (Director,

Cyberinfrastructure Enables Full Solutions to Einstein’s Equations

100 year quest: EE’s can’t be solved since 1916 Decades of struggle in community

Individual PIs, groups, grand challenges, communities 2002-2005: Breakthoughs for 2BH orbits: general

EEs can now be solved by many groups!Now ready for the Universe!

Brownsville Pretorius* AEI/LSU Goddard Waveforms!

Page 5: Changing Science and Engineering: the impact of HPC Sept 23, 2009 Edward Seidel Assistant Director, Mathematical and Physical Sciences, NSF (Director,

Two decades later: Complexity of UniverseLHC, Gamma-ray bursts!

Gamma-ray bursts!• All energy emitted in lifetime of sun

bursts out in a few seconds: what are they?! Colliding BH-NS? SN?

• GR, hydrodynamics, nuclear physics, radiation transport, neutrinos, magnetic fields: globally distributed collab!

• Scalable algorithms, complex AMR codes, viz, PFlops*week, PB output!

LHC: What is the nature of mass? Higgs particle?

• ~10K scientists, 33+ countries, 25PB data, distributed!

• Planetary lab for scientific discovery!

Now, compare with observation...LIGO, GEO, VIRGO! GR becomes a collaborative data-driven science

Remote InstrumentRemote Instrument

Remote InstrumentRemote Instrument

Page 6: Changing Science and Engineering: the impact of HPC Sept 23, 2009 Edward Seidel Assistant Director, Mathematical and Physical Sciences, NSF (Director,

Data-Driven Multiscale Collaborations* for ComplexityGreat Challenges of 21st Century

HPC began the revolution, but it continues in 4 dimensions!Multiscale Collaborations

• General Relativity, Particles, Geosciences, Bio, Social...

• Data need to be shared in collab... Science and Society being

transformed by CI and Data• Compute, data, network volumes

grow 9-12 orders of magnitude in two decades

• Completely new methodologies• HPC previewed fundamental shift to

data-driven science: will be different

Science needs place requirements on data access

*Small groups still important!*Small groups still important!

Page 7: Changing Science and Engineering: the impact of HPC Sept 23, 2009 Edward Seidel Assistant Director, Mathematical and Physical Sciences, NSF (Director,

Summary

After 4 centuries of constancy in science methods and culture, we have experienced a profound step function in change9-12 orders of magnitude in data, compute, network,

and exponentially growingCollaborations growing dramatically in all areas as we

solve complex challenges HPC was just the beginning

Data-driven science computational methods (including HPC) are the foreseeable future in every discipline

Data are the media of collaboration