Hybrid-view programming of nuclear fusion simulation code in the PGAS parallel programming language XcalableMP

Abstract

Recently, the Partitioned Global Address Space (PGAS) parallel programming model has emerged as a usable distributed memory programming model. Xcal-ableMP (XMP) is a PGAS parallel programming language that extends base languages such as C and Fortran with directives in OpenMP-like style. XMP supports a global-view model that allows programmers to define global data and to map them to a set of processors, which execute the distributed global data as a single thread. In XMP, the concept of a coarray is also employed for local-view programming. In this study, we port Gyrokinetic Toroidal Code-Princeton (GTC-P), which is a three-dimensional gyrokinetic PIC code developed at Princeton University to study the microturbulence phenomenon in magnetically confined fusion plasmas, to XMP as an example of hybrid memory model coding with the global-view and local-view programming models. In local-view programming, the coarray notation is simple and intuitive compared with Message Passing Interface (MPI) programming while the performance is comparable to that of the MPI version. Thus, because the global-view programming model is suitable for expressing the data parallelism for a field of grid space data, we implement a hybrid-view version using a global-view programming model to compute the field and a local-view programming model to compute the movement of particles. The performance is degraded by 20% compared with the original MPI version, but the hybrid-view version facilitates more natural data expression for static grid space data (in the global-view model) and

Performance evaluation for a hydrodynamics application in XcalableACC PGAS language for accelerated clusters

“…For applications with complex and undescribable decomposition and communication in the directives, the XACC local-view model's coarray feature in the accelerator memory allows us to describe complex communication with simple syntax. Moreover, combining the global-view and local-view models is possible in XACC, such as the XMP-hybrid view model [17]. As mentioned earlier, the XACC version still requires the addition of code and modifications compared with the serial version without directives and performance degradation may occur due to unnecessary updates in the halo exchange process.…”

“…Some application codes have been implemented with XMP. Thus, Tsugane et al implemented and evaluated a nuclear fusion simulation code with XMP [17], where they proposed a hybrid-view model that used both a global-view model based on directives and a local-view model based on the coarray feature [13] in the application. The global-view model was used for static grid space data and the local-view mode was used for dynamic particle data.…”