James R. Geraci
ABSTRACT
Our work focuses on the area of using a high level language to improve program productivity, performance and portability. In general, this has been an area of intense research. There are a number of previous efforts including ZPL [Chamberlain and et al 2004], X10/Fortress/Chapel from IBM/SUN/Cray [Weiland 2007], Intel's CT/RapidMind [McCool 2006] and parallel VSIPL++ [Lebak and et al 2005] to name a few. However, while these languages do great things in simplifying parallel implementation of code, extensions beyond that are limited. The primary exception to this is VSIPL++ which implements several high level functions useful to the signal processing community. While most of these languages can be used to implement graphics or game related algorithms if necessary, none of them attempt to provide a platform that makes such development particularly easy. On the other hand, high level engines such as Renderman and Unreal provide the wanted abstractions but with little or no guarantees about extensibility, portability, or parallel performance. Our research focuses on adapting the parallel VSIPL++ API from the signal processing community to the graphics and game development environment.
- Chamberlain, B. L., and et al. 2004. The high-level parallel language ZPL improves productivity and performance. Proceedings of the IEEE International Workshop on Productivity and Performance in High-End Computing.Google Scholar
- Lebak, J., and et al. 2005. Parallel VSIPL++: An open standard software library for high-performance parallel signal processing. Proceedings of the IEEE 93, 2 (February), 313--330.Google ScholarCross Ref
- McCool, M. D. 2006. Data-parallel programming on the Cell BE and the GPU using the RapidMind development platform. GSPx Multicore Applications Conference, Santa Clara.Google Scholar
- Weiland, M. 2007. Chapel, Fortress and X10: novel languages for HPC. Tech. rep., EPCC, The University of Edinburgh, October.Google Scholar
Index Terms
- Improving program productivity, performance and portability through a high level language for graphics and game development
Recommendations
High-Productivity Software Development with pMatlab
pMatlab, a parallel Matlab library, lowers the barrier to development of parallel and distributed application codes by providing an easy-to-use programming environment.
Productivity and Performance of Global-View Programming with XcalableMP PGAS Language
CCGRID '12: Proceedings of the 2012 12th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing (ccgrid 2012)XcalableMP (XMP) is a PGAS parallel language with a directive-based extension of C and Fortran. While it supports gcoarray as a local-view programming model, an XMP global-view programming model is useful when parallelizing data-parallel programs by ...
Productivity and performance using partitioned global address space languages
PASCO '07: Proceedings of the 2007 international workshop on Parallel symbolic computationPartitioned Global Address Space (PGAS) languages combine the programming convenience of shared memory with the locality and performance control of message passing. One such language, Unified Parallel C (UPC) is an extension of ISO C defined by a ...
Comments