arXiv subjects

Vladimir Loncar

Publications and source records attributed to Vladimir Loncar.

39 records · Page 3Linked to original sources

arXiv2020

Fast inference of Boosted Decision Trees in FPGAs for particle physics

We describe the implementation of Boosted Decision Trees in the hls4ml library, which allows the translation of a trained model into FPGA firmware through an automated conversion process. Thanks to its fully on-chip implementation, hls4ml performs inference of Boosted Decision Tree models with extremely low latency. With a typical latency less than 100 ns, this solution is suitable for FPGA-based real-time processing, such as in the Level-1 Trigger system of a collider experiment. These developments open up prospects for physicists to deploy BDTs in FPGAs for identifying the origin of jets, better reconstructing the energies of muons, and enabling better selection of rare signal processes.

physics.comp-ph↗

arXiv2019

C and Fortran OpenMP programs for rotating Bose-Einstein condensates

We present OpenMP versions of C and Fortran programs for solving the Gross-Pitaevskii equation for a rotating trapped Bose-Einstein condensate (BEC) in two (2D) and three (3D) spatial dimensions. The programs can be used to generate vortex lattices and study dynamics of rotating BECs. We use the split-step Crank-Nicolson algorithm for imaginary- and real-time propagation to calculate stationary states and BEC dynamics, respectively. The programs propagate the condensate wave function and calculate several relevant physical quantities, such as the energy, the chemical potential, and the root-mean-square sizes. The imaginary-time propagation starts with an analytic wave function with one vortex at the trap center, modulated by a random phase at different space points. Nevertheless, the converged wave function for a rapidly rotating BEC with a large number of vortices is most efficiently calculated using the pre-calculated converged wave function of a rotating BEC containing a smaller number of vortices as the initial state rather than using an analytic wave function with one vortex as the initial state. These pre-calculated initial states exhibit rapid convergence for fast-rotating condensates to states containing multiple vortices with an appropriate phase structure. This is illustrated here by calculating vortex lattices with up to 61 vortices in 2D and 3D. Outputs of the programs include calculated physical quantities, as well as the wave function and different density profiles (full density, integrated densities in lower dimensions, and density cross-sections). The provided real-time propagation programs can be used to study the dynamics of a rotating BEC using the imaginary-time stationary wave function as the initial state. We also study the efficiency of parallelization of the present OpenMP C and Fortran programs with different compilers.

cond-mat.quant-gas↗

arXiv2017

OpenMP GNU and Intel Fortran programs for solving the time-dependent Gross-Pitaevskii equation

We present Open Multi-Processing (OpenMP) version of Fortran 90 programs for solving the Gross-Pitaevskii (GP) equation for a Bose-Einstein condensate in one, two, and three spatial dimensions, optimized for use with GNU and Intel compilers. We use the split-step Crank-Nicolson algorithm for imaginary- and real-time propagation, which enables efficient calculation of stationary and non-stationary solutions, respectively. The present OpenMP programs are designed for computers with multi-core processors and optimized for compiling with both commercially-licensed Intel Fortran and popular free open-source GNU Fortran compiler. The programs are easy to use and are elaborated with helpful comments for the users. All input parameters are listed at the beginning of each program. Different output files provide physical quantities such as energy, chemical potential, root-mean-square sizes, densities, etc. We also present speedup test results for new versions of the programs.

physics.comp-ph↗