Performance Evaluation and Optimization on GPU

Abstract:

GPU provides higher peak performance with hundreds of cores than CPU counterpart. However, it is a big challenge to take full advantage of their computing power. In order to understand performance bottlenecks of applications on many-core GPU and then optimize parallel programs on GPU architectures, we propose a performance evaluating model based on memory wall and then classify applications into AbM (Application bound-in Memory) and AbC (Application bound-in Computing). Furthermore, we optimize kernels characterized with low memory bandwidth including matrix multiplication and FFT (Fast Fourier Transform) by employing texture cache on NVIDIA GTX280 using CUDA (Compute Unified Device Architecture). Experimental results show that texture cache is helpful for AbM with better data locality, so it is critical to utilize GPU memory hierarchy efficiently for performance improvement.

• Periodical: Advanced Materials Research (Volumes 219-220)
• Pages: 1445-1449
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.219-220.1445
• Online since: March 2011
• Authors: Xin Biao Gan, Li Shen, Quan Yuan Tan, Cong Liu, Zhi Ying Wang
• Keywords: Application Bound-In Memory, CUDA, Optimization, Performance Evaluating Model

References

[1] Ryoo S., Rodrigues, C. I., Baghsorkhi, S. S., etc. Optimization principles and application performance evaluation of a multithreaded GPU using CUDA. In Proceedings of the 13th ACM SIGPLAN Symposium on Principles and practice of parallel programming, Salt Lake. 2008, pp.73-82.

DOI: 10.1145/1345206.1345220

[2] NVIDIA. CUDA programming guide 2.0. NVIDIA Corporation, (2008).

[3] NVIDIA GeForce series GTX280, 8800GTX, 8800GT. http://www.nvidia.com/geforce.

[4] Nathan B., Michael G. Implementing sparse matrix-vector multiplication on throughput oriented processors. In Proceedings of the ACM SC09, New York.2009, pp.141-152.

[5] Ogata, Y., Endo T., Maruyama N.etc. An efficient, model-based CPU-GPU heterogeneous FFT library. In Proceedings of the 22nd IEEE International Parallel and Distributed Processing Symposium, Piscataway. 2008, pp.252-257.

DOI: 10.1109/ipdps.2008.4536163

[6] Khailany B., Dally W. J., Rixner S., etc. Imagine: Media Processing with Streams. IEEE Micro, 2001, p.35–46.

DOI: 10.1109/40.918001

[7] Dally W. J, Labonte, F., Das A., etc. Merrimac: Supercomputing with Streams. In Proceedings of the SC2003. Nov. 15-21, Phoenix, Arizona, (2003).

[8] Eichenberger A. E., Brien K. O., Brien, K. O., etc. Optimizing Compiler for the CELL Processor. In Proceedings of the PACT2005. Washington, DC, USA, 2005, pp.161-172.