This Article 
 Bibliographic References 
 Add to: 
EcoG: A Power-Efficient GPU Cluster Architecture for Scientific Computing
March/April 2011 (vol. 13 no. 2)
pp. 83-87

Researchers built the EcoG GPU-based cluster to show that a system can be designed around GPU computing and still be power efficient.

1. Tesla C2050 and Tesla C2070 Computing Processor Board Specification, Nvidia, 2010; .
2. G. Shi et al., "Design of MILC Lattice QCD Application for GPU Clusters," Proc. IEEE Int'l Parallel and Distributed Processing Symp. (IPDPS), IEEE CS Press, 2011, accepted.
3. T. Chen et al., "Cell Broadband Engine Architecture and Its First Implementation—A Performance View," IBM J. Research and Development, vol. 51, no. 5, 2007, pp. 559–572.
4. R. Stevens, The LLNL/AND/IBM Collaboration to Develop GB/P and BG/Q, Argonne Nat'l Lab, 2010; Mar06Stevens.pdf.
5. J. Enos et al., "Quantifying the Impact of GPUs on Performance and Energy Efficiency in HPC Clusters," Proc. Int'l Conf. Green Computing, IEEE CS Press, 2010;

Index Terms:
Graphics processing, GPUs, Nvidia, CUDA, scientific computing
Mike Showerman, Jeremy Enos, Craig Steffen, Sean Treichler, William Gropp, Wen-mei W. Hwu, "EcoG: A Power-Efficient GPU Cluster Architecture for Scientific Computing," Computing in Science and Engineering, vol. 13, no. 2, pp. 83-87, March-April 2011, doi:10.1109/MCSE.2011.30
Usage of this product signifies your acceptance of the Terms of Use.