Efficient Computational Workload Distribution on Heterogeneous GPUs

Abstract:

Recently, heterogeneous system architectures are becoming a mainstream for achieving high performance and power efficiency. In particular, many-core graphics processing units (GPUs) have started to play an important role for computing in heterogeneous architectures. However, for application designers, computational workload still needs to be distributed among heterogeneous GPUs manually and remains inefficient. In this work, we propose a MINLP-based method for efficient workload distribution among GPUs by considering the capabilities of GPUs for various applications. Experimental results demonstrate the performance of our proposed method.

• Periodical: Applied Mechanics and Materials (Volumes 479-480)
• Pages: 805-809
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.479-480.805
• Online since: December 2013
• Authors: Chih Sheng Lin, Po Ting Liu, Chih Wei Hsieh, Hsi Ya Chang, Pao Ann Hsiung
• Keywords: Computational Workload Distribution, Graphic Processing Units (GPUs), Load Balancing, Mixed-Integer Non-Linear Programming (MINLP), Program Configuration

References

[1] Top500. http: /www. top500. org.

[2] Nvidia tegra. http: /www. nvidia. com/object/tegra. html.

[3] NVIDIA Corporation. NVIDIA CUDA Programming Guide. (2009).

[4] J.E. Stone, D. Gohara, and G. Shi. OpenCL: A parallel programming standard for heterogeneous computing systems. Computing in Science and Engineering, 12(3): 66. (2010).

DOI: 10.1109/mcse.2010.69

[5] Microsoft. DirectCompute. http: /www. microsoft. com/en-us/download/details. aspx?id=27731.

[6] C. -K. Luk, S. Hong, and H. Kim. Qilin: exploiting parallelism on heterogeneous multiprocessors with adaptive mapping. In the Proceedings of 42nd Annual IEEE/ACM International Symposium on Microarchitecture. MICRO-42, pages 45–55. (2009).

DOI: 10.1145/1669112.1669121

[7] W. Liu, Z. Du, Y. Xiao, D.A. Bader, and C. Xu. A waterfall model to achieve energy efficient tasks mapping for large scale gpu clusters. In the Proceedings of IEEE International Symposium on Parallel and Distributed Processing Workshops and Phd Forum (IPDPSW), pages 82–92. (2011).

DOI: 10.1109/ipdps.2011.129

[8] Al´ecio P.D. Binotto, C.E. Pereira, and D.W. Fellner. Towards dynamic reconfigurable load-balancing for hybrid desktop platforms. In the Proceedings of IEEE International Symposium on Parallel & Distributed Processing Workshops and Phd Forum (IPDPSW), pages 1–4. (2010).

DOI: 10.1109/ipdpsw.2010.5470804

[9] I. Galindo, F. Almeida, and J.M. Bad´ıa-Contelles. Dynamic load balancing on dedicated heterogeneous systems. In Recent Advances in Parallel Virtual Machine and Message Passing Interface, Springer, pages 64–74. (2008).

DOI: 10.1007/978-3-540-87475-1_14

[10] S. Che, M. Boyer, J. Meng, D. Tarjan, J. W. Sheaffer, S. -H. Lee, and K. Skadron. Rodinia: A benchmark suite for heterogeneous computing. In the Proceedings of IEEE International Symposium on Workload Characterization (IISWC), pages 44–54. (2009).

DOI: 10.1109/iiswc.2009.5306797

[11] Nvidia, GPU computing SDK. https: /developer. nvidia. com/gpu-computing-sdk.