Anshu Dubey, together with colleagues from Turkey, Switzerland, France and the United States, is co-organizer of the fifth workshop on Programming Abstractions for Data Locality (PADAL’19), to be held Sept. 9–11, 2019, in Bordeaux, France.
Data locality — moving computation close to where the data are, rather than moving large amounts of data over the network — has become a key concern in high-performance computing (HPC). In response, researchers have developed abstractions that can expose crucial information about data locality to the compiler and runtime system.
“Data locality abstractions are now available in libraries, languages and runtime systems and are enabling performance portability of codes,” said Dubey, a computer scientist in Argonne’s Mathematics and Computer Science division. “But emerging computer architectures raise particular challenges, and we anticipate lively discussions of these issues,” Dubey said.
The PADAL’19 workshop will focus on extreme heterogeneity while managing data locality on diverse accelerators. Topics include locality abstractions for different types of memory technologies and data locality concerns beyond Moore’s law. Locality concerns for both HPC and non-HPC applications on accelerators will also be explored.
More information about the workshop is available here.