It is expected that the first exascale supercomputer will be deployed within the next 10 years, but the programming model which allows easy development and high performance is still unknown. APGAS programming model offers a flexible way for wide range of applications to express many patterns of concurrency, communication, and control through the combination of asynchronous operations and a global view of data. However, the comparative performance of APGAS model with
the existing standard message passing model such as MPI remains unclear. In this work, we give a detailed comparative analysis of APGAS model in X10 with the standard message passing model, by using lattice Quantum Chromodynamics (QCD) as an example, which is one of the most challenging applications for supercomputers. We further analyze the performance of lattice QCD in X10 and apply several optimizations. Our experimental results show that our X10 implementation scales up to 256 places. The results also show that the MPI-based implementation performs 2.26x--2.58x faster than the X10 implementation.
各種検索
サポート
ヘルプ