Naoya Maruyama

I am a Team Leader at RIKEN Advanced Institute for Computational Science (RIKEN AICS) , where I lead the HPC Programming Framework Research Team. I joined RIKEN AICS in 2012 after years at Tokyo Institute of Technology, where I received Ph.D. in Computer Science in 2008.

Research Interests

Current Projects

My current primary focus of research is high-level parallel frameworks for computational science applications to support productive and highly efficient computing with large-scale parallel systems such as RIKEN’s K computer. One of the software deliverables that have been developed under that vision is a high-performance implementation of MapReduce for large-scale supercomputers and clusters, K MapReduce (KMR). The MapReduce programming model is an abstraction particularly suited for inherently parallel data processing. Our KMR provides a simple library-based implemenation of MapReduce optimized for HPC-centric parallel systems, where high-speed networking is available with Message Passing Interface. KMR can also exploit the hierarchical storage system of the K computer, which is our primary target system. See the project page for more information.

I am also the Principal Investigator for a JST Post Petascale System Software Project, where our team has been developing domain-specific programming models for mesh-based and particle-based applications that are primarily targeted to achieve high scalability and efficiency on large-scale GPU machines, such as Physis.

Some open positions may be available; see the RIKEN Career page if interested in joining our team (look for “HPC Programming Framework Research Team”).

Recent Papers and Talks

See the (almost) up-to-date list here. Also, here and here.