John Robert Wernsing
Abstract
Over the past decade, system architectures have started on a clear trend towards increased parallelism and heterogeneity, often resulting in speedups of 10x to 100x. Despite numerous compiler and high-level synthesis studies, usage of such systems has largely been limited to device experts, due to significantly increased application design complexity. To reduce application design complexity, we introduce elastic computing - a framework that separates functionality from implementation details by enabling designers to use specialized functions, called elastic functions, which enable an optimization framework to explore thousands of possible implementations, even ones using different algorithms. Elastic functions allow designers to execute the same application code efficiently on potentially any architecture and for different runtime parameters such as input size, battery life, etc. In this paper, we present an initial elastic computing framework that transparently optimizes application code onto diverse systems, achieving significant speedups ranging from 1.3x to 46x on a hyper-threaded Xeon system with an FPGA accelerator, a 16-CPU Opteron system, and a quad-core Xeon system.
- J. Ansel, C. Chan, Y.L. Wong, M. Olszewskim, Q. Zhao, A. Edelman, and S. Amarasinghe. PetaBricks: A Language and Compiler for Algorithmic Choice. Proceedings of the ACM SIGPLAN Conference on Programming Language Design and Implementation (PLDI), 2009, pp. 38--49. Google Scholar
Digital Library
- B. Chamberlain, D. Callahan, and H. Zima. Parallel Programmability and the Chapel Language. International Journal of High Performance Computing Applications, Vol. 21, Issue 3, August 2007, pg. 291--312. Google ScholarDigital Library
- W. Chen, D. Bonachea, J. Duell, P. Husbands, C. Iancu, and K. Yelick. A Performance Analysis of the Berkeley UPC Compiler. Proceedings of the International Conference on Supercomputing (ICS), 2003, pg. 63--73. Google ScholarDigital Library
- Cray, Inc. Cray XT5 System. 2008. http://www.cray.com/Products/XT/Product/Technology.aspx.Google Scholar
- A. DeHon. The Density Advantage of Configurable Computing. Computer, Vol. 33, Issue 4, April 2000, pp 41--49. Google ScholarDigital Library
- ElementCXI, Inc. ECA-64. http://www.elementcxi.com/productbrief.html.Google Scholar
- A. Fin, F. Fummi, and M. Signoretto. SystemC: A Homogenous Environment to Test Embedded Systems. Proceedings of the International Workshop on Hardware/Software Codesign (CODES), 2001, pp 17--22. Google ScholarDigital Library
- M. Frigo and S. Johnson. FFTW: an Adaptive Software Architecture for the FFT. Acoustics, Speech and Signal Processing. Proceedings of the IEEE International Conference on Acoustics, Speech, and Signal Processing, 1998, pp. 1381--1384.Google Scholar
- M. Girkar and C. Polychronopoulos. Extracting Task-Level Parallelism. ACM Transactions on Programming Languages and Systems (TOPLAS), Vol. 17, Issue 4, July 1995, pp. 600--634. Google ScholarDigital Library
- B. Grattan, G. Stitt and F. Vahid. Codesign-Extended Applications. IEEE/ACM International Symposium on Hardware/Software Codesign (CODES), 2002, pp. 1--6. Google ScholarDigital Library
- Z. Guo, W. Najjar, F. Vahid, and K. Vissers. A Quantitative Analysis of the Speedup Factors of FPGAs over Processors. Proceedings of the International Symposium on Field Programmable Gate Arrays (FPGA), pp. 162--170, 2004. Google ScholarDigital Library
- S. Gupta, N. Dutt, R. Gupta, and A. Nicolau. SPARK: A High-Level Synthesis Framework for Applying Parallelizing Compiler Transformations. Proceedings of International Conference on VLSI Design (VLSI), 2003. Google ScholarDigital Library
- H. Peter Hofstee. Power Efficient Processor Architecture and the Cell Processor. Proceedings of the International Symposium on High Performance Computer Architecture (HPCA), 2005, pg. 258--262. Google ScholarDigital Library
- B. Holland, K. Nagarajan, C. Conger, A. Jacobs, and A. George. RAT: a Methodology for Predicting Performance in Application Design Migration to FPGAs. Proceedings of the Workshop on High-Performance Reconfigurable Computing Technology and Applications (HPRCTA), pp 1--10, 2007. Google ScholarDigital Library
- Intel Quad-Core Xeon. 2008. http://www.intel.com.Google Scholar
- L. Lewins and K. Prager. Experience and Results Porting HPEC Benchmarks to MONARCH. Proceedings of Workshop on High Performance Embedded Computing (HPEC), 2008.Google Scholar
- C. Luk, S. Hong, and H. Kim. Qilin: Exploiting Parallelism on Heterogeneous Multiprocessors with Adaptive Mapping. Proceedings of the IEEE/ACM International Symposium on Microarchitecture (MICRO), 2009, pg. 45--55. Google ScholarDigital Library
- M. Macedonia. The GPU Enters Computing's Mainstream. IEEE Computer, Vol. 36, No. 10, October 2003, pp. 106--108. Google ScholarDigital Library
- I. McCallum. Intel QuickAssist Technology Accelerator Abstraction Layer (AAL) 317481-001US. 2007. http://download.intel.com/technology/platforms/quickassist/quickassist_aal_whitepaper.pdf.Google Scholar
- M. D. McCool. Data-parallel programming on Cell BE and the GPU using the Rapidmind development platform. In GSPx Multicore Applications Conference, 2006.Google Scholar
- S. Merchant, B. Holland, C. Reardon, et al. Strategic Challenges for Application Development Productivity in Reconfigurable Computing. Proceedings of the IEEE National Areospace and Electronics Conference (NAECON), 2008.Google ScholarCross Ref
- K. Morris. FPGAs in Space: Programmable Logic in Orbit. FPGA and Structured ASIC Journal, August, 2004.Google Scholar
- D. Musser. Introspective Sorting and Selection Algorithms. Software: Practice and Experience, Vol. 27, Issue 8, 1999, pp. 983--993. Google ScholarDigital Library
- Nallatech Inc. Nallatech PCIXM FPGA accelerator card, 2008. http://www.nallatech.com/?node_id=1.2.2&id=41.Google Scholar
- G. R. Nudd, D. J. Kerbyson, E. Papaefstathiou, S. C. Perry, J. S. Harper, and D. V. Wilcox. Pace - A Toolset for the Performance Prediction of Parallel and Distributed Systems. International Journal of High Performance Computing Applications, Vol. 14, No. 3, 2000, pp. 228--251. Google ScholarDigital Library
- L. Semeria, K. Sato, and G. De Micheli. Synthesis of Hardware Models in C with Pointers and Complex Data Structures. IEEE Transactions of Very Large Scale Integration Systems (TVLSI), Vol. 9, Issue 6, December 2001, pp. 743--756. Google ScholarDigital Library
- G. Stitt, F. Vahid, and W. Najjar. A Code Refinement Methodology for Performance-Improved Synthesis from C. Proceedings of the IEEE/ACM International Conference on Computer-Aided Design (ICCAD), 2006, pp. 716--723 Google ScholarDigital Library
- Tilera Tile64 Processor Family. 2008. http://www.tilera.com/products/processors.php.Google Scholar
- R. Vuduc, J. Demmel, and K. Yelick. OSKI: A Library of Automatically Tuned Sparse Matrix Kernels. Journal of Physics, June 2005.Google Scholar
- R. Whaley and J. Dongarra. Automatically Tuned Linear Algebra Software. Proceedings of ACM/IEEE Conference on Supercomputing (SC), 1998, pp. 1--27. Google ScholarDigital Library
- J. Williams, A. George, J. Richardson, K. Gosrani, and S. Suresh. Fixed and Reconfigurable Multi-Core Device Characterization for HPEC. Proceedings of Workshop on High-Performance Embedded Computing (HPEC), 2008.Google Scholar
- Xilinx Inc. Virtex IV FX devices, 2008. http://www.xilinx.com/products/silicon_solutions/fpgas/virtex/virtex4/index.htm.Google Scholar
Index Terms
- Elastic computing: a framework for transparent, portable, and adaptive multi-core heterogeneous computing
Recommendations
Elastic computing: a framework for transparent, portable, and adaptive multi-core heterogeneous computing
LCTES '10: Proceedings of the ACM SIGPLAN/SIGBED 2010 conference on Languages, compilers, and tools for embedded systemsOver the past decade, system architectures have started on a clear trend towards increased parallelism and heterogeneity, often resulting in speedups of 10x to 100x. Despite numerous compiler and high-level synthesis studies, usage of such systems has ...
Elastic computing: A portable optimization framework for hybrid computers
Due to power limitations and escalating cooling costs, high-performance computing systems can no longer rely solely on faster clock frequencies and numerous microprocessor nodes to meet increasing performance demands. As an alternative approach, high-...
RACECAR: a heuristic for automatic function specialization on multi-core heterogeneous systems
PPoPP '12: Proceedings of the 17th ACM SIGPLAN symposium on Principles and Practice of Parallel ProgrammingHigh-performance computing systems increasingly combine multi-core processors and heterogeneous resources such as graphics-processing units and field-programmable gate arrays. However, significant application design complexity for such systems has often ...
Comments