Abstract
An existing hybrid MPI-OpenMP scheme is augmented with a CUDA-based fine grain parallelization approach for multidimensional distributed Fourier transforms, in a well-characterized pseudospectral fluid turbulence code. Basics of the hybrid scheme are reviewed, and heuristics provided to show a potential benefit of the CUDA implementation. The method draws heavily on the CUDA runtime library to handle memory management and on the cuFFT library for computing local FFTs. The manner in which the interfaces to these libraries are constructed, and ISO bindings utilized to facilitate platform portability, are discussed. CUDA streams are implemented to overlap data transfer with cuFFT computation. Testing with a baseline solver demonstrated significant aggregate speed-up over the hybrid MPI-OpenMP solver by offloading to GPUs on an NVLink-based test system. While the batch streamed approach provided little benefit with NVLink, we saw a performance gain of 30 % when tuned for the optimal number of streams on a PCIe-based system. It was found that strong GPU scaling is nearly ideal, in all cases. Profiling of the CUDA kernels shows that the transform computation achieves 15% of the attainable peak FlOp-rate based on a roofline model for the system. In addition to speed-up measurements for the fiducial solver, we also considered several other solvers with different numbers of transform operations and found that aggregate speed-ups are nearly constant for all solvers.
Highlights
Turbulence in the atmosphere and in the oceans prevails over a very wide range of scales, from planetary motions down to the millimeter scale where dissipation begins to set in Reference [1,2].The nonlinearities characteristic of turbulence lead to the development in the flow of intermittency, in the form of strong localized events and multi-fractal scaling [3,4]
Heuristics tailored for our application are used to indicate the way in which transfer to and from the device can affect the gains otherwise achieved by reducing computational cost on the GPUs, and we show how CUDA streams may be used to diminish the impact of data transfer on different systems
While the motivation in M11 was mainly to show the overall efficacy of this hybrid MPI-OpenMP parallelization scheme, we described the specific directives for thread parallelization and the cache-blocking procedure used for the local transpose step
Summary
The nonlinearities characteristic of turbulence lead to the development in the flow of intermittency, in the form of strong localized events and multi-fractal scaling [3,4]. These extreme events impact the dynamics of atmospheric and oceanic flows, such as, e.g., in the observed persistence of strong aerosol contents in the atmosphere [5]. Flows in the atmosphere and in the oceans often display dynamics at disparate time scales, with slow wave motions as those observed to be persistent over Antarctica at large scales [6], fast waves, such as as gravity waves, which are often modeled in current global. Understanding the interplay between such waves and turbulent eddies is a key question in the study of geophysical flows, which is in many cases answered using highly accurate numerical simulations
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
