SCIENTIFIC COMPUTING & VISUALIZATION
Lecture Notes
- Introduction;
slides
- Basic molecular dynamics algorithm;
linked-list cell MD algorithm;
slides
- Message Passing Interface;
slides;
getting started with Discovery cluster
- Parallel computation of Pi--scalability
- Parallel MD algorithm;
introduction;
slides;
in situ analysis of MD simulation data using communicators;
Shaw's NT algorithm;
see
A fast, scalable method for the parallel evaluation of distance-limited pairwise particle interactions,
D. E. Shaw, J. Comput. Chem 26, 1318 (2005),
A scalable parallel algorithm for dynamic range-limited n-tuple computation in many-body molecular dynamics simulation,
M. Kunaseth et al., Proc. of Supercomputing, SC13 (ACM/IEEE, 2013) and
Shift/collapse on neighbor list (SC-NBL): fast evaluation of dynamic many-body potentials in molecular dynamics simulations,
M. Kunaseth et al., Comput. Phys. Commun. 235, 88 (2019)
- OpenMP
- Hybrid MPI+OpenMP parallel MD;
see
Performance characteristics of hardware transactional memory for molecular dynamics application on Blue Gene/Q:
toward efficient multithreading strategies for large-scale scientific applications,
M. Kunaseth et al., best paper of Proc. of PDSEC13 (IEEE, 2013);
MPI+X,
M. Wolfe, HPC Wire (2014);
MPI+MPI,
T. Hoefler et al., Computing 95, 1121 (2013).
- Visualizing molecular dynamics;
slides;
how to install OpenGL and GLUT
- VMD & OVITO visualization of molecular dynamics;
see Parallel in situ visualization with a fully featured visualization system,
B. Whitlock et al., Proc. of PGV11 (Eurographics, 2011) and
Biological electron transfer pathway analysis plugin for VMD,
I. A. Balabin et al., J. Comput. Chem. 33, 906 (2012);
A framework for stochastic simulations and visualization of biological electron-transfer dynamics,
C. M. Nakano et al., Comput. Phys. Commun. 193, 1 (2015) and
an associated movie
- Massive dataset visualization; see
A. Sharma et al.,
Immersive and interactive exploration of billion-atom systems,
Presence 12, 85 ('03),
X. Liu et al.,
Improving interactivity of a parallel and distributed immersive walkthrough application
for very large datasets with artificial neural network-based machine learning,
Proc. PDPTA IV, 2054 ('02),
C. Zhang et al.,
ParaViz: a spatially decomposed parallel visualization algorithm using hierarchical visibility ordering,
Int'l J. Comput. Sci. 1, 407 ('07);
for tensor-field visualization, see:
L. Hesselink et al.,
Research issues in vector and tensor field visualization,
IEEE Computer Graphics & Applications 14, 76 ('93),
W. Ribarsky et al.,
Glyphmaker: creating customized visualizations of complex data,
IEEE Computer 27(7), 57 ('94),
A. Sigfridsson et al.,
Tensor field visualisation using adaptive filtering of noise fields combined with glyph rendering,
IEEE Visualization 2002, (IEEE, 2002) p. 371,
E. Zhang et al.,
Interactive tensor field design and visualization on surfaces,
IEEE T. Vis. Comput. Graph. 13, 94 ('07)
- Virtual-reality programming;
Game-Engine-Assisted Research platform for Scientific computing (GEARS) in virtual reality,
B. Horton et al., SoftwareX 9, 112 (2019)
- CUDA; see
L. Peng et al., Parallel lattice Boltzmann flow simulation on emerging multi-core platforms,
Proc. Euro-Par 2008, 763 (2008);
P. E. Small et al., Acceleration of dynamic n-tuple computations in many-body molecular dynamics,
Proc. HPC Asia 2018 (2018);
Sasan Tavakkol's final project became a poster
in GPU Technology Conference
(see nice videos 1 & 2);
C. Rizzo et al., PAR2: parallel random walk particle tracking method for solute transport in porous media,
Comput. Phys. Commun. 239, 265 (2019);
J. C. Phillips et al., Scalable molecular dynamics on CPU and GPU architectures with NAMD,
J. Chem. Phys. 153, 044130 (2020);
S. Pall et al., Heterogeneous parallelization and acceleration of molecular dynamics simulations in GROMACS,
J. Chem. Phys. 153, 134110 (2020)
- Pair distribution computation with CUDA;
see also B. G. Levine et al.,
J. Comput. Phys. 230, 3556 (2011)
- Hybrid MPI+OpenMP+CUDA computing
- See HPC Asia presentation by Jack Dongarra on reduced-precision computation & tensor cores.
- OpenMP target offload
- OpenMP 4.5 specification
- SYCL
- DPC++ book
by B. Ashbaugh et al. (Intel, 2020, unedited review copy)
- Optimizing molecular dynamics; see Berkeley CS267 lecture on
Single processor machines: memory hierarchies and processor features by Prof. Kathy Yelick;
for BLAS-ification, see K. Nomura et al.,
Metascalable quantum molecular dynamics simulations of hydrogen-on-demand,
Prof. Supercomputing, SC14 (IEEE/ACM, '14); for MD optimization, see
J. Mellor-Crummey et al.,
Improving memory hierarchy performance for irregular applications using data and computation reorderings,
Int'l J. Par. Prog. 29, 217 ('01);
Intel Vtune performance amplifier
- Advanced topics in parallel molecular dynamics
- Metascalable divide-conquer-recombine algorithmic framework;
Quantum molecular dynamics in the post-petaflop/s era,
N. A. Romero et al., IEEE Computer 48(11), 33-41 (2015)
- Grid computing
- MapReduce;
using Hadoop at USC-HPC
- Quantum computing hands-on:
lecture on quantum computing for science;
hands-on exercise on
qubits and quantum gates and
quantum dynamics simulation
- GitHub
- Miscellaneous lectures: