Figure 1.
Data layout of buffer used for transfer of the Wigner coupling coefficients and quantum numbers of irreducible tensors computed on CPU to GPU
-
Previous Article
Segmentation of color images using mean curvature flow and parametric curves
- DCDS-S Home
- This Issue
-
Next Article
Existence of a period two solution of a delay differential equation
March
2021, 14(3): 1111-1122.
doi: 10.3934/dcdss.2020383
Transformation of a Nucleon-Nucleon potential operator into its SU(3) tensor form using GPUs
| 1. | Faculty of Nuclear Sciences and Physical Engineering, Czech Technical University in Prague, Trojanova 13, Praha 2,120 00, Czech Republic |
| 2. | Department of Physics and Astronomy, Louisiana State University, Baton Rouge, LA 70803, USA, Nuclear Physics Institute, Czech Academy of Sciences, Řež 25068, Czech Republic |
| 3. | Department of Physics and Astronomy, Louisiana State University, Baton Rouge, LA 70803, USA |
| 4. | Faculty of Information Technology, Czech Technical University, Prague 16000, Czech Republic, Aerospace Research and Test Establishment, Prague 19905, Czech Republic |
| 5. | Department of Physics and Astronomy, Louisiana State University, Baton Rouge, LA 70803, USA |
Starting from the matrix elements of a nucleon-nucleon potential operator provided in a basis of spherical harmonic oscillator functions, we present an algorithm for expressing a given potential operator in terms of irreducible tensors of the SU(3) and SU(2) groups. Further, we introduce a GPU-based implementation of the latter and investigate its performance compared with a CPU-based version of the same. We find that the CUDA implementation delivers speedups of 2.27x – 5.93x.
Mathematics Subject Classification:
Primary: 81V35, 81R15; Secondary: 68W10.
Citation:
Tomáš Oberhuber, Tomáš Dytrych, Kristina D. Launey, Daniel Langr, Jerry P. Draayer. Transformation of a Nucleon-Nucleon potential operator into its SU(3) tensor form using GPUs. Discrete & Continuous Dynamical Systems - S,
2021, 14
(3)
: 1111-1122.
doi: 10.3934/dcdss.2020383
![]()
References:
| [1] |
Y. Akiyama and J. P. Draayer, A user's guide to Fortran programs for Wigner and Racah coefficients of SU$_3$, Comp. Phys. Comm, 5 (1973), 405-406. Google Scholar |
| [2] |
T. Dytrych, K. D. Launey, J. P. Draayer, P. Maris, J. P. Vary, E. Saule, U. Catalyurek, M. Sosonkina, D. Langr and M. A. Caprio, Collective modes in light nuclei from first principles, Phys. Rev. Lett., 111 (2013), 252501.
doi: 10.1103/PhysRevLett.111.252501. |
| [3] |
T. Dytrych, P. Maris, K. D. Launey, J. P. Draayer, J. Vary, D. Langr, E. Saule, M. A. Caprio, U. Catalyurek and M. Sosonkina,
Efficacy of the SU(3) scheme for ab initio large-scale calculations beyond the lightest nuclei, Comp. Phys. Comm., 207 (2016), 202-210.
doi: 10.2172/1326837. |
| [4] |
H. T. Johansson and C. Forssén, Fast and accurate evaluation of Wigner $3j$, $6j$, and $9j$ symbols using prime factorization and multiword integer arithmetic, SIAM J. Sci. Comput., 38 (2016), A376–A384.
doi: 10.1137/15M1021908. |
| [5] |
K. D. Launey, T. Dytrych and J. P. Draayer,
Symmetry-guided large-scale shell-model theory, Prog. Part. Nucl. Phys., 89 (2016), 101-136.
doi: 10.1016/j.ppnp.2016.02.001. |
| [6] |
M. F. O'Reilly,
A closed formula for the product of irreducible representations of SU(3), J. Math. Phys., 23 (1982), 2022-2028.
doi: 10.1063/1.525258. |
show all references
References:
| [1] |
Y. Akiyama and J. P. Draayer, A user's guide to Fortran programs for Wigner and Racah coefficients of SU$_3$, Comp. Phys. Comm, 5 (1973), 405-406. Google Scholar |
| [2] |
T. Dytrych, K. D. Launey, J. P. Draayer, P. Maris, J. P. Vary, E. Saule, U. Catalyurek, M. Sosonkina, D. Langr and M. A. Caprio, Collective modes in light nuclei from first principles, Phys. Rev. Lett., 111 (2013), 252501.
doi: 10.1103/PhysRevLett.111.252501. |
| [3] |
T. Dytrych, P. Maris, K. D. Launey, J. P. Draayer, J. Vary, D. Langr, E. Saule, M. A. Caprio, U. Catalyurek and M. Sosonkina,
Efficacy of the SU(3) scheme for ab initio large-scale calculations beyond the lightest nuclei, Comp. Phys. Comm., 207 (2016), 202-210.
doi: 10.2172/1326837. |
| [4] |
H. T. Johansson and C. Forssén, Fast and accurate evaluation of Wigner $3j$, $6j$, and $9j$ symbols using prime factorization and multiword integer arithmetic, SIAM J. Sci. Comput., 38 (2016), A376–A384.
doi: 10.1137/15M1021908. |
| [5] |
K. D. Launey, T. Dytrych and J. P. Draayer,
Symmetry-guided large-scale shell-model theory, Prog. Part. Nucl. Phys., 89 (2016), 101-136.
doi: 10.1016/j.ppnp.2016.02.001. |
| [6] |
M. F. O'Reilly,
A closed formula for the product of irreducible representations of SU(3), J. Math. Phys., 23 (1982), 2022-2028.
doi: 10.1063/1.525258. |
Figure 2.
Data layout of buffer holding records with triplets of references to records in buffer $ B_W $ . Each such triplet thus represents one irreducible tensor
Table 1.
Performance results obtained on the Blue Waters system. Notation $ +1 $ in the column with MPI processes denotes one master process which just assigns HO shells to other MPI processes. The master process is not taken into account for efficiency evaluation
| MPI procs. | CPU only | CPU+GPU | |||
| Time [s] | Efficiency | Time [s] | Speed-up | ||
| 7+1 | 295.3 | – | 75.3 | 3.92 | |
| 15+1 | 137.6 | 1 | 36.8 | 3.73 | |
| 31+1 | 66.5 | 1 | 17.5 | 3.79 | |
| 63+1 | 39.4 | 0.83 | 9.23 | 4.26 | |
| 127+1 | 32.8 | 0.56 | 6.35 | 5.17 | |
| 255+1 | 31.0 | 0.52 | 5.22 | 5.94 | |
| 7+1 | 2219 | – | 648 | 3.42 | |
| 15+1 | 1034 | 1 | 318 | 3.24 | |
| 31+1 | 499 | 1 | 151 | 3.28 | |
| 63+1 | 248 | 0.99 | 74 | 3.32 | |
| 127+1 | 165 | 0.74 | 43 | 3.75 | |
| 255+1 | 138 | 0.59 | 32 | 4.25 | |
| 7+1 | 13083 | – | 4493 | 2.91 | |
| 15+1 | 6097 | 1 | 2116 | 2.88 | |
| 31+1 | 2943 | 1 | 1054 | 2.79 | |
| 63+1 | 1447 | 1 | 515 | 2.80 | |
| 127+1 | 776 | 0.92 | 269 | 2.88 | |
| 255+1 | 565 | 0.68 | 169 | 3.33 | |
| 5+1 | 64865 | – | 26104 | 2.48 | |
| 15+1 | 30227 | 1 | 12204 | 2.47 | |
| 31+1 | 14596 | 1 | 5944 | 2.45 | |
| 63+1 | 7179 | 1 | 2932 | 2.44 | |
| 127+1 | 3581 | 0.99 | 1461 | 2.45 | |
| 255+1 | 2142 | 0.83 | 838 | 2.55 | |
| MPI procs. | CPU only | CPU+GPU | |||
| Time [s] | Efficiency | Time [s] | Speed-up | ||
| 7+1 | 295.3 | – | 75.3 | 3.92 | |
| 15+1 | 137.6 | 1 | 36.8 | 3.73 | |
| 31+1 | 66.5 | 1 | 17.5 | 3.79 | |
| 63+1 | 39.4 | 0.83 | 9.23 | 4.26 | |
| 127+1 | 32.8 | 0.56 | 6.35 | 5.17 | |
| 255+1 | 31.0 | 0.52 | 5.22 | 5.94 | |
| 7+1 | 2219 | – | 648 | 3.42 | |
| 15+1 | 1034 | 1 | 318 | 3.24 | |
| 31+1 | 499 | 1 | 151 | 3.28 | |
| 63+1 | 248 | 0.99 | 74 | 3.32 | |
| 127+1 | 165 | 0.74 | 43 | 3.75 | |
| 255+1 | 138 | 0.59 | 32 | 4.25 | |
| 7+1 | 13083 | – | 4493 | 2.91 | |
| 15+1 | 6097 | 1 | 2116 | 2.88 | |
| 31+1 | 2943 | 1 | 1054 | 2.79 | |
| 63+1 | 1447 | 1 | 515 | 2.80 | |
| 127+1 | 776 | 0.92 | 269 | 2.88 | |
| 255+1 | 565 | 0.68 | 169 | 3.33 | |
| 5+1 | 64865 | – | 26104 | 2.48 | |
| 15+1 | 30227 | 1 | 12204 | 2.47 | |
| 31+1 | 14596 | 1 | 5944 | 2.45 | |
| 63+1 | 7179 | 1 | 2932 | 2.44 | |
| 127+1 | 3581 | 0.99 | 1461 | 2.45 | |
| 255+1 | 2142 | 0.83 | 838 | 2.55 | |
Table 2.
Performance results obtained for 32 MPI processes running on two 16-core IBM Power9 CPUs and a single NVIDIA V100 GPU
| CPU only | CPU+GPU | ||
| Time [s] | Time [s] | Speed-up | |
| 8 | 41.65 | 15.78 | 2.63 |
| 10 | 274.14 | 97.99 | 2.79 |
| 12 | 1649.7 | 611.1 | 2.69 |
| 14 | 7761.9 | 3407.6 | 2.27 |
| CPU only | CPU+GPU | ||
| Time [s] | Time [s] | Speed-up | |
| 8 | 41.65 | 15.78 | 2.63 |
| 10 | 274.14 | 97.99 | 2.79 |
| 12 | 1649.7 | 611.1 | 2.69 |
| 14 | 7761.9 | 3407.6 | 2.27 |
| [1] |
Liqun Qi, Shenglong Hu, Yanwei Xu. Spectral norm and nuclear norm of a third order tensor. Journal of Industrial & Management Optimization, 2021 doi: 10.3934/jimo.2021010 |
| [2] |
Jelena Rupčić. Convergence of lacunary SU(1, 1)-valued trigonometric products. Communications on Pure & Applied Analysis, 2020, 19 (3) : 1275-1289. doi: 10.3934/cpaa.2020062 |
| [3] |
Jan J. Sławianowski, Vasyl Kovalchuk, Agnieszka Martens, Barbara Gołubowska, Ewa E. Rożko. Essential nonlinearity implied by symmetry group. Problems of affine invariance in mechanics and physics. Discrete & Continuous Dynamical Systems - B, 2012, 17 (2) : 699-733. doi: 10.3934/dcdsb.2012.17.699 |
| [4] |
Elie Assémat, Marc Lapert, Dominique Sugny, Steffen J. Glaser. On the application of geometric optimal control theory to Nuclear Magnetic Resonance. Mathematical Control & Related Fields, 2013, 3 (4) : 375-396. doi: 10.3934/mcrf.2013.3.375 |
| [5] |
Bertrand Lods. Variational characterizations of the effective multiplication factor of a nuclear reactor core. Kinetic & Related Models, 2009, 2 (2) : 307-331. doi: 10.3934/krm.2009.2.307 |
| [6] |
Florian Caro, Bilal Saad, Mazen Saad. Study of degenerate parabolic system modeling the hydrogen displacement in a nuclear waste repository. Discrete & Continuous Dynamical Systems - S, 2014, 7 (2) : 191-205. doi: 10.3934/dcdss.2014.7.191 |
| [7] |
Erlend Grong, Alexander Vasil’ev. Sub-Riemannian and sub-Lorentzian geometry on $SU(1,1)$ and on its universal cover. Journal of Geometric Mechanics, 2011, 3 (2) : 225-260. doi: 10.3934/jgm.2011.3.225 |
| [8] |
James Benn. Fredholm properties of the $L^{2}$ exponential map on the symplectomorphism group. Journal of Geometric Mechanics, 2016, 8 (1) : 1-12. doi: 10.3934/jgm.2016.8.1 |
| [9] |
Wen Li, Wei-Hui Liu, Seak Weng Vong. Perron vector analysis for irreducible nonnegative tensors and its applications. Journal of Industrial & Management Optimization, 2021, 17 (1) : 29-50. doi: 10.3934/jimo.2019097 |
| [10] |
E. N. Dancer. On domain perturbation for super-linear Neumann problems and a question of Y. Lou, W-M Ni and L. Su. Discrete & Continuous Dynamical Systems, 2012, 32 (11) : 3861-3869. doi: 10.3934/dcds.2012.32.3861 |
| [11] |
Duo Wang, Zheng-Fen Jin, Youlin Shang. A penalty decomposition method for nuclear norm minimization with l1 norm fidelity term. Evolution Equations & Control Theory, 2019, 8 (4) : 695-708. doi: 10.3934/eect.2019034 |
| [12] |
Bernard Bonnard, Monique Chyba, Alain Jacquemard, John Marriott. Algebraic geometric classification of the singular flow in the contrast imaging problem in nuclear magnetic resonance. Mathematical Control & Related Fields, 2013, 3 (4) : 397-432. doi: 10.3934/mcrf.2013.3.397 |
| [13] |
Rasmus Backholm, Tatiana A. Bubba, Camille Bélanger-Champagne, Tapio Helin, Peter Dendooven, Samuli Siltanen. Simultaneous reconstruction of emission and attenuation in passive gamma emission tomography of spent nuclear fuel. Inverse Problems & Imaging, 2020, 14 (2) : 317-337. doi: 10.3934/ipi.2020014 |
| [14] |
Inmaculada Antón, Julián López-Gómez. Global bifurcation diagrams of steady-states for a parabolic model related to a nuclear engineering problem. Conference Publications, 2013, 2013 (special) : 21-30. doi: 10.3934/proc.2013.2013.21 |
| [15] |
J. F. Padial. Existence and estimate of the location of the free-boundary for a non local inverse elliptic-parabolic problem arising in nuclear fusion. Conference Publications, 2011, 2011 (Special) : 1176-1185. doi: 10.3934/proc.2011.2011.1176 |
| [16] |
Maria Bortos, Joe Gildea, Abidin Kaya, Adrian Korban, Alexander Tylyshchak. New self-dual codes of length 68 from a $ 2 \times 2 $ block matrix construction and group rings. Advances in Mathematics of Communications, 2020 doi: 10.3934/amc.2020111 |
| [17] |
Joe Gildea, Abidin Kaya, Adam Michael Roberts, Rhian Taylor, Alexander Tylyshchak. New self-dual codes from $ 2 \times 2 $ block circulant matrices, group rings and neighbours of neighbours. Advances in Mathematics of Communications, 2021 doi: 10.3934/amc.2021039 |
| [18] |
Sergio Estrada, J. R. García-Rozas, Justo Peralta, E. Sánchez-García. Group convolutional codes. Advances in Mathematics of Communications, 2008, 2 (1) : 83-94. doi: 10.3934/amc.2008.2.83 |
| [19] |
Heping Liu, Yu Liu. Refinable functions on the Heisenberg group. Communications on Pure & Applied Analysis, 2007, 6 (3) : 775-787. doi: 10.3934/cpaa.2007.6.775 |
| [20] |
Stefan Haller, Tomasz Rybicki, Josef Teichmann. Smooth perfectness for the group of diffeomorphisms. Journal of Geometric Mechanics, 2013, 5 (3) : 281-294. doi: 10.3934/jgm.2013.5.281 |
2020 Impact Factor: 2.425
Tools
Metrics
Other articles
by authors
[Back to Top]