Abstract of GOMIC Project

Predicting the performance of fusion plasmas in terms of amplification factor, namely the ratio of the fusion power over the injected power, is among the key challenges in fusion plasma physics. In this perspective, turbulence and heat transport need being modeled within the most accurate theoretical framework, using first-principle nonlinear simulation tools.

Using the nonlinear global full-f gyrokinetic 5D code GYSELA, a simulation very close to ITER-size plasmas with kinetic ions and adiabatic electrons has been performed on 16384 cores during 15 days. This type of simulation is at the front-edge of current research in fusion plasma modeling.

In each part of the code GYSELA, parallel algorithms have been designed in order to scale up to thousands of cores using a hybrid MPI/OpenMP approach. Each year, the development of new methods and techniques are conceived in order to remove bottlenecks concerning memory scalability, communication costs, and outputs on disks. Recently, we have achieved an efficiency of 91% on the full Juqueen machine composed of 458,752 cores [1].

Intel MIC architecture is steadily being adopted in clusters. The current generation MIC coprocessor, Xeon Phi, provides a highly multi-threaded environment. Regular programming models such as MPI/OpenMP have started utilizing systems with MIC coprocessors. This specific hardware offers large memory bandwidth compared to standard computes nodes. As GYSELA is partly bound by memory bandwidth, ways to overcome this limit have to be considered, and MIC device is one of those ways.

[1] http://www.fz-juelich.de/ias/jsc/EN/Expertise/High-Q-Club/_node.html