Gaspar data-centric framework

This paper presents the Gaspar data-centric framework to develop high performance parallel applications in Java. Our approach is based on data iterators and on the map pattern of computation. The framework provides an efficient data Application Programming Inter-face(API) that supports flexible data layout and data tiling. Data layout and tiling enable the improvement of data locality, which is essential to foster application scalability in modern multi-core systems. The paper presents the framework data-centric concepts and shows that the performance is comparable to pure Java code.(undefined)