Title: Synthetic Programming Using CorePy

Speaker: Chris Mueller and Andrew Lumsdaine (Indiana University)

Date/Time: Monday, August 27, 2007, 10:30 am - 11:30 am

Location: IPB in Research Park, Room 1149 (Sandia NM)

Brief Abstract: Scripting and interpreted languages are important tools for software engineering and are often used in place of compiled languages for application development.  While they enable a high level of developer productivity, their run-time environments limit the overall performance attainable with any given application. To develop performance-critical applications, developers continue to rely on compiled languages. However, tuning applications in compiled languages is a time consuming process that affects the overall productivity of development teams. Additionally, recent changes in microprocessor design have begun to push the limits of compiled languages, making it difficult for developers to fully utilize the available hardware resources.

In this talk we present CorePy, a Python package for interactive development of multi-core and assembly-level applications on IBM Cell/B.E. and PowerPC-based processors.  CorePy is based synthetic programming, a technique for run-time code synthesis from scripting languages that lets developers build and execute optimized programs without resorting to compiled languages.  CorePy includes APIs for managing asynchronous execution on multi-core processors from Python and provides complete access to all features of the PowerPC, VMX, and SPU instruction sets, giving developers fine-grained control over optimization strategies. CorePy has been used to develop applications on PowerPC-based systems, IBM Cell Blades, and Sony Playstation 3 consoles.