Faisal Saied
back to main page
INCITE proposal prep pdf
Running hpl
ITaP Lunch Series, August 22, 2006
Important reports on High Performance Computing:
1982:
Lax report:
Large Scale Computing in Science and Engineering
1993:
Branscomb report:
From desktop to Teraflop: Exploiting the U.S. lead in High Performance Computing
1995:
Hayes report:
Future of the NSF Supercomputing Centers Program
2003:
Atkins report:
Revolutionizing Science and Engineering through Cyberinfrastructure
2005:
NRC:
Getting up to Speed: The Future of Supercomputing
2005:
PITAC:
Computational Science: Ensuring America's Competitiveness
NSF Investing in America's Future. Strategic Plan 2006 - 2011
Jan 20, 2006:
NSF's Cyberinfrastructure Vision
for 21st Century Discovery
Report of the High Performance Computing Town Hall Meeting,
Science, Requirements, and Benchmarks.
October 17, 2005, University of Illinois at Urbana-Champaign.
Ingredients of an Advanced Cyberinfrastructure
A. High Performance Computing (Path to the Petaflop)
B. Data, Data Analysis & Visualization
C. Collaboration, Communication & Remote Access
D. Education & Training
NSF Goals:
Creating a World-Class HPC Environment
to enable Petascale Science and Engineering
See also
PPT slides from NSF's Office of Cyberinfrastructure (OCI).
DoE, Office of Science, SciDAC
SciDAC-2: The Next Phase of Discovery
In the Petascale RFP
the National
Science Foundations poses the following questions, and states that HPC can
help researchers answer them.
1. What are the three-dimensional structures of all of the proteins encoded by
an organism's genome and how does structure influence function, both
spatially and temporally?
2. What patterns of emergent behavior occur in
models of very large societies?
3. How do massive stars explode and produce
the heaviest elements in the periodic table?
4. What sort of abrupt
transitions can occur in Earth's climate and ecosystem structure? How
do these occur and under what circumstances?
5. If we could design catalysts
atom-by-atom, could we transform industrial synthesis?
6. What strategies
might be developed to optimize management of complex infrastructure
systems?
7. What kind of language processing can occur in large assemblages of
neurons?
8. Can we enable integrated planning and response to natural and
man-made disasters that prevent or minimize the loss of life and property?
Some applications areas that need HPC
Climate Modeling
Weather Forecasting
Storm Modeling
Computational Nanotechnology
Astrophysics
Cosmology
Numerical Relativity
High Energy Physics
Quantum Chromodynamics
Condensed Matter Physics
Computational Fluid Dynamics
Numerical Wind Tunnel; Aircraft design
Molecular Dynamics
Proteomics
BioInformatics
Ion Channel Simulations
Virus Structure
Bio-medical Informatics
Bio-medical Engineering
Drug Design
Geophysics
Seismic Modeling
Oil Reservoir Simulations
Earthquake modeling, engineering
Groundwater modeling
Aeronautical engineering
Computational Chemistry
Crash testing (automotive industry)
Nuclear Engineering
Centers
RCAC
TACC
NCSA
NERSC
SDSC
OSC
RENCI, UNC
SI, Mn
CCT, LSU
OSCER
Architectures
IBM BlueGene/L
Mare Nostrum (IBM JS20)
Bigred (IBM JS21)
IBM p690
Earth Simulator
Cray XT3
Cray X1
SGI Altix
Xeon Linux cluster
Opteron cluster
Advanced Processor Architectures
IBM
Power6
PowerPC
Cell
Intel
64 bit Xeon
Itanium 2
Woodcrest
Multi-core
AMD
Opteron
Interconnection Technologies
Mricom
Infiniband
10 Gigabit Ethernet
Gigabit Ethernet
Parallel Applications/Community Codes
MILC
LAMMPS
Cactus
CCSM
Enzo
Gaussian
WRF
LS-DYNA
NAMD
sPPM
ZEUS-MP
Data Intensive Computing
Biology Workbench
Blast
CMS, Tier 2
National Virtual Observatory
Sloan Digital Sky Survey
BIRN
Data-mining
Parallel Programming
MPI
OpenMPI
OpenMP
P-Threads
Parallel I/O
Co-array Fortran
Unified Parallel C
TotalView
Guide
Parallel Performance Analysis
TAU
PAPI
MPIP
MPE, Jumpshot
Vampir
Paraver
Dimemas
svPABLO
HPM
VTune
Intel Cluster Tools
(MPI, OpenMP, Intel Trace Analyzer and Collector for Linux, MKL)
Ganglia
Useful Compiler Optimizations
Loop unrolling
Pre-fetching
Software pipeling
Loop fission
Loop fusion
Branch prediction
Interprocedural Anaysis
Inlining
Vectorization
Architecture specific compilation
Risky Optimizations:
Relax IEEE 754 compliance
re-order floating point operations
Alter rounding; no gradual underflow
Data dependency analysis
Latency hiding
Profile-guided optimizations
OpenMP directives
Automatic parallelization
Benchmarks
NSF benchmarks
SPEC HPC
Top 500
Kazushge Goto
HPC Challenge
Stream
NetPIPE
Pallas
NAS Parallel Benchmark
NCSA Bench
ONRL Cray X1 Evaluation
Parallel Scientific Libraries
PETSc
FFTw
ScaLAPACK
PARPACK
NAG
Hypre
SuperLU
Computational Problems; Algorithms
Sparse Matrix Methods
Krylov Subspace Solvers
Multiscale Methods
Multigrid
Algebraic Multigrid
Domain Decomposition
Lanczos, Arnoldi
Trace Minimization
Jacobi-Davidson
Preconditioners
Sparse Direct Solvers
SVD
FFT
Wavelet transform
Fast Multipole Method
Fast Poisson Solvers
Spectral Methods
Finite Element Method
Mesh Generation
Graph partitioning
Automatic Differentiation
Nonlinear Solvers, Newton-Krylov-Schwarz
Time-stepping methods
Nonlinear least squares solvers
Optimization Methods
Scientific Visualization
Cactus
Galaxy
LBL Vignettes
Rossmann: cold virus
Mike Norman: Orion Nebula
UIUC, CSAR
NERSC: Fuel Cells
NERSC: Galaxy simulation
NERSC: Photosynthesis
NERSC: Richtmyer-Meshkov instability
NERSC: SuperLU, Tokamak
NERSC: Gamma ray burst 1
NERSC: Gamma ray burst 2
NERSC: Fusion waves
NERSC: Mis-folding
NERSC: CdSe Nanocrystals
PSC: Schulten-1
PSC: Schulten-2
SchultenL Aquaporin
Science Portals
Nanohub
Biology Workbench
GridChem
NCBI
GridPort
NCSA Alliance Science Portal
List of Science Portals
Grid Computing
Northwest Indiana Computational Grid
Teragrid
Open Science Grid
UK E-Science
Condor
Optiputer
LHC Computing Grid
seti@home
Globus
IBM On demand computing