SmartApps, an Application Centric Approach to High Performance Computing: Compiler-Assisted Software and Hardware Support for Reduction Operations

State-of-the-art run-time systems are a poor match to di-verse, dynamic distributed applications because they are designed to provide support to a wide variety of applica-tions, without much customization to individual specific re-quirements. Little or no guiding information flows directly from the application to the run-time system to allow the lat-ter to fully tailor its services to the application. As a result, the performance is disappointing. To address this prob-lem, we propose application-centric computing, or SMART APPLICATIONS. In the executable of smart applications, the compiler embeds most run-time system services, and a performance-optimizing feedback loop that monitors the application’s performance and adaptively reconfigures the application and the OS/hardware platform. At run-time, after incorporating the code’s input and the system’s re-sources and state, the SMARTAPP performs a global op-timization. This optimization is instance specific and thus much more tractable than a global generic optimization be-tween application, OS and hardware. The resulting code and resource customization should lead to major speedups. In this paper, we first describe the overall architecture of SMARTAPPS and then present some achievements to date, focusing on compiler-assisted software and hardware tech-niques for parallelizing reduction operations. These illus-trate SMARTAPPS use of adaptive algorithm selection and moderately reconfigurable hardware.