Scalability of 3D-DFT by block tensor-matrix multiplication on the JUWELS Cluster
Authors:
Nitin Malapally,
Viacheslav Bolnykh,
Estela Suarez,
Paolo Carloni,
Thomas Lippert,
Davide Mandelli
Abstract:
The 3D Discrete Fourier Transform (DFT) is a technique used to solve problems in disparate fields. Nowadays, the commonly adopted implementation of the 3D-DFT is derived from the Fast Fourier Transform (FFT) algorithm. However, evidence indicates that the distributed memory 3D-FFT algorithm does not scale well due to its use of all-to-all communication. Here, building on the work of Sedukhin \text…
▽ More
The 3D Discrete Fourier Transform (DFT) is a technique used to solve problems in disparate fields. Nowadays, the commonly adopted implementation of the 3D-DFT is derived from the Fast Fourier Transform (FFT) algorithm. However, evidence indicates that the distributed memory 3D-FFT algorithm does not scale well due to its use of all-to-all communication. Here, building on the work of Sedukhin \textit{et al}. [Proceedings of the 30th International Conference on Computers and Their Applications, CATA 2015 pp. 193-200 (01 2015)], we revisit the possibility of improving the scaling of the 3D-DFT by using an alternative approach that uses point-to-point communication, albeit at a higher arithmetic complexity. The new algorithm exploits tensor-matrix multiplications on a volumetrically decomposed domain via three specially adapted variants of Cannon's algorithm. It has here been implemented as a C++ library called S3DFT and tested on the JUWELS Cluster at the Jülich Supercomputing Center. Our implementation of the shared memory tensor-matrix multiplication attained 88\% of the theoretical single node peak performance. One variant of the distributed memory tensor-matrix multiplication shows excellent scaling, while the other two show poorer performance, which can be attributed to their intrinsic communication patterns. A comparison of S3DFT with the Intel MKL and FFTW3 libraries indicates that currently iMKL performs best overall, followed in order by FFTW3 and S3DFT. This picture might change with further improvements of the algorithm and/or when running on clusters that use network connections with higher latency, e.g. on cloud platforms.
△ Less
Submitted 23 March, 2023;
originally announced March 2023.
Wavefunction-based electrostatic-embedding QM/MM using CFOUR through MiMiC
Authors:
Till Kirsch,
Jógvan Magnus Haugaard Olsen,
Viacheslav Bolnykh,
Simone Meloni,
Emiliano Ippoliti,
Ursula Rothlisberger,
Michele Cascella,
Jürgen Gauss
Abstract:
We present an interface of the wavefunction-based quantum-chemical software CFOUR to the multiscale modeling framework MiMiC. Electrostatic embedding of the quantummechanical (QM) part is achieved by analytic evaluation of one-electron integrals in CFOUR, while the rest of the QM/MM operations are treated according to the previous MiMiC-based QM/MM implementation. Long-range electrostatic interact…
▽ More
We present an interface of the wavefunction-based quantum-chemical software CFOUR to the multiscale modeling framework MiMiC. Electrostatic embedding of the quantummechanical (QM) part is achieved by analytic evaluation of one-electron integrals in CFOUR, while the rest of the QM/MM operations are treated according to the previous MiMiC-based QM/MM implementation. Long-range electrostatic interactions are treated by a multipole expansion of the potential from the QM electron density to reduce the computational cost without loss of accuracy. Testing on model water/water systems, we verified that the CFOUR interface to MiMiC is robust, guaranteeing fast convergence of the SCF cycles and optimal conservation of the energy during the integration of the equations of motion. Finally, we verified that the CFOUR interface to MiMiC is compatible with the use of a QM/QM multiple time-step algorithm, which effectively reduces the cost of AIMD or QM/MM-MD simulations using higher level wavefunction-based approaches compared to cheaper density-functional theory based ones. The new wavefunction-based AIMD and QM/MM-MD implementation was tested and validated for a large number of wavefunction approaches, including Hartree-Fock and post-Hartree-Fock methods like Moller-Plesset, coupled cluster, and complete active space self-consistent field.
△ Less
Submitted 24 August, 2021;
originally announced August 2021.
