Search

OpenQ$^\star$D code has been used by the RC$^\star$ collaboration for the generation of fully dynamical QCD+QED gauge configurations with C$^\star$ boundary conditions. In this talk, optimization of solvers provided with the openQ$^\star$D package relevant for porting the code on GPU-accelerated supercomputing platforms is discussed. We present the analysis of the current implementations of the GCR solver preconditioned with Schwarz alternating procedure for ill-conditioned Dirac-operators. With the goal of enabling support for GPUs from various vendors, a novel method of adaptive CPU/GPU-hybrid implementation is proposed.

To elucidate the geometric structure of the putative room temperature superconductor, carbonaceous sulfur hydride, at high pressure, we present the results of an extensive computational structure search of bulk C-S-H at 250 gigapascals. Using the minima hopping structure prediction method coupled to the GPU accelerated Sirius library, more than 17,000 local minima with different stochiometries in large simulation cells were investigated. Only 24 stochiometries are favourable against elemental decomposition, all of them are carbon doped H$_3$S crystals. The absence of van Hove singularities or similar peaks in the electronic density of states of more than 3.000 candidate phases rules out conventional superconductivity in C-S-H at room-temperature.

Matrix factorizations are among the most important building blocks of scientific computing. State-of-the-art libraries, however, are not communication-optimal, underutilizing current parallel architectures. We present novel algorithms for Cholesky and LU factorizations that utilize an asymptotically communication-optimal 2.5D decomposition. We first establish a theoretical framework for deriving parallel I/O lower bounds for linear algebra kernels, and then utilize its insights to derive Cholesky and LU schedules, both communicating N^3/(P*sqrt(M)) elements per processor, where M is the local memory size. The empirical results match our theoretical analysis: our implementations communicate significantly less than Intel MKL, SLATE, and the asymptotically communication-optimal CANDMC and CAPITAL libraries. Our code outperforms these state-of-the-art libraries in almost all tested scenarios, with matrix sizes ranging from 2,048 to 262,144 on up to 512 CPU nodes of the Piz Daint supercomputer, decreasing the time-to-solution by up to three times. Our code is ScaLAPACK-compatible and available as an open-source library., Comment: 15 pages (including references), 11 figures. arXiv admin note: substantial text overlap with arXiv:2010.05975

Communication-avoiding algorithms for Linear Algebra have become increasingly popular, in particular for distributed memory architectures. In practice, these algorithms assume that the data is already distributed in a specific way, thus making data reshuffling a key to use them. For performance reasons, a straightforward all-to-all exchange must be avoided. Here, we show that process relabeling (i.e. permuting processes in the final layout) can be used to obtain communication optimality for data reshuffling, and that it can be efficiently found by solving a Linear Assignment Problem (Maximum Weight Bipartite Perfect Matching). Based on this, we have developed a Communication-Optimal Shuffle and Transpose Algorithm (COSTA): this highly-optimised algorithm implements $A=\alpha\cdot \operatorname{op}(B) + \beta \cdot A,\ \operatorname{op} \in \{\operatorname{transpose}, \operatorname{conjugate-transpose}, \operatorname{identity}\}$ on distributed systems, where $A, B$ are matrices with potentially different (distributed) layouts and $\alpha, \beta$ are scalars. COSTA can take advantage of the communication-optimal process relabeling even for heterogeneous network topologies, where latency and bandwidth differ among nodes. The implementation not only outperforms the best available ScaLAPACK redistribute and transpose routines multiple times, but is also able to deal with more general matrix layouts, in particular it is not limited to block-cyclic layouts. Finally, we use COSTA to integrate a communication-optimal matrix multiplication algorithm into the CP2K quantum chemistry simulation package. This way, we show that COSTA can be used to unlock the full potential of recent Linear Algebra algorithms in applications by facilitating interoperability between algorithms with a wide range of data layouts, in addition to bringing significant redistribution speedups., Comment: To be published in the proceedings of the 36th International Conference on High Performance Computing, ISC High Performance 2021. The implementation of the algorithm is available at: https://github.com/eth-cscs/COSTA

We report APW+${lo}$ (augmented plane wave plus local orbital) density functional theory (DFT) calculations of molecule systems using the domain specific SIRIUS multi-functional DFT package. Compared to other packages the additional APW and FLAPW task and data parallelism and the additional eigensystem solver provided by the SIRIUS package can be exploited for performance gains in in the ground state Kohn-Sham calculation. This is in contrast with the use of SIRIUS as a library backend to some other APW+${lo}$ or FLAPW (full-potential linearized AWP) code. We benchmark the code and demonstrate performance on several magnetic molecule and metal organic framework systems. We show that the SIRIUS package in itself is capable of handling systems as large as a few hundreds of atoms in the unit cell without losing the accuracy needed for magnetic systems., Comment: 41 pages, 6 figures

To address ultimate precision in density-functional-theory calculations we employ the full-potential linearized augmented planewave + local-orbital (LAPW+lo) method and justify its usage as a benchmark method. LAPW+lo and two completely unrelated numerical approaches, multi-resolution analysis (MRA) and linear combination of atomic orbitals, yield total energies of atoms with a mean deviation of 0.9~{\mu}Ha and 0.2~{\mu}Ha, respectively. Spectacular agreement with the MRA is reached also for total and atomization energies of the G2-1 set consisting of 55 molecules. With the example of $\alpha$-iron we demonstrate the capability of LAPW+lo of reaching {\mu}Ha/atom precision also for periodic systems, which allows also for distinction between numerical precision and the accuracy of a given functional., Comment: 6 pages, 3 figures

We report complex band structure (CBS) calculations for the four late transition metal monoxides, MnO, FeO, CoO and NiO, in their paramagnetic phase. The CBS is obtained from density functional theory plus dynamical mean field theory (DMFT) calculations to take into account correlation effects. The so-called $\beta$ parameters, governing the exponential decay of the transmission probability in the non-resonant tunneling regime of these oxides, are extracted from the CBS. Different model constructions are examined in the DMFT part of the calculation. The calculated $\beta$ parameters provide theoretical estimation for the decay length in the evanescent channel, which would be useful for tunnel junction applications of these materials., Comment: 16 pages, 13 figures

The GW approximation is a well-known method to improve electronic structure predictions calculated within density functional theory. In this work, we have implemented a computationally efficient GW approach that calculates central properties within the Matsubara-time domain using the modified version of Elk, the full-potential linearized augmented plane wave (FP-LAPW) package. Continuous-pole expansion (CPE), a recently proposed analytic continuation method, has been incorporated and compared to the widely used Pade approximation. Full crystal symmetry has been employed for computational speedup. We have applied our approach to 18 well-studied semiconductors/insulators that cover a wide range of band gaps computed at the levels of single-shot G0W0, partially self-consistent GW0, and fully self-consistent GW (scGW). Our calculations show that G0W0 leads to band gaps that agree well with experiment for the case of simple s-p electron systems, whereas scGW is required for improving the band gaps in 3-d electron systems. In addition, GW0 almost always predicts larger band gap values compared to scGW, likely due to the substantial underestimation of screening effects. Both the CPE method and Pade approximation lead to similar band gaps for most systems except strontium titantate, suggesting further investigation into the latter approximation is necessary for strongly correlated systems. Our computed band gaps serve as important benchmarks for the accuracy of the Matsubara-time GW approach., Comment: 4 tables, 3 figures

We present a new algorithm to analytically continue the self-energy of quantum many-body systems from Matsubara frequencies to the real axis. The method allows straightforward, unambiguous computation of electronic spectra for lattice models of strongly correlated systems from self-energy data that has been collected with state-of-the are continuous time solvers within dynamical mean field simulations. Using well-known analytical properties of the self-energy, the analytic continuation is cast into a constrained minimization problem that can be formulated as a quadratic programmable optimization with linear constraints. The algorithm is validated against exactly solvable finite size problems, showing that all features of the spectral function near the Femi level are very well reproduced and coarse features are reproduced for all energies. The method is applied to two well known lattice problems, the two-dimensional Hubbard model at half filling where the momentum dependence of the gap formation is studied, as well as a multi-band model of NiO, for which the spectral function can be directly compared to experiment. Agreement with results published results is very good.

Layered 5d transition metal oxides (TMOs) have attracted significant interest in recent years because of the rich physical properties induced by the interplay between spin-orbit coupling, bandwidth and on-site Coulomb repulsion. In Sr2IrO4, this interplay opens a gap near the Fermi energy and stabilizes a Jeff=1/2 spin-orbital entangled insulating state at low temperatures. Whether this metal-insulating transition (MIT) is Mott-type (electronic-correlation driven) or Slater-type (magnetic-order driven) has been under intense debate. We address this issue via spatially resolved imaging and spectroscopic studies of the Sr2IrO4 surface using scanning tunneling microscopy/spectroscopy (STM/S). The STS results clearly illustrate the opening of the (~150-250 meV) insulating gap at low temperatures, in qualitative agreement with our density-functional theory (DFT) calculations. More importantly, the measured temperature dependence of the gap width coupled with our DFT+dynamical mean field theory (DMFT) results strongly support the Slater-type MIT scenario in Sr2IrO4. The STS data further reveal a pseudogap structure above the Neel temperature, presumably related to the presence of antiferromagnetic fluctuations.