Type of Document Master's Thesis Author Lee, Kenneth Sydney Author's Email Address KLee1@vt.edu URN etd-07272012-152625 Title Characterization and Exploitation of GPU Memory Systems Degree Master of Science Department Computer Science and Applications Advisory Committee
Advisor Name Title Feng, Wu-Chun Committee Chair Cao, Yong Committee Member Lin, Heshan Committee Member Keywords
- Data Transfer
- Performance Modeling
- Memory Systems
Date of Defense 2012-07-06 Availability unrestricted AbstractGraphics Processing Units (GPUs) are workhorses of modern performance due to their ability to achieve massive speedups on parallel applications. The massive number of threads that can be run concurrently on these systems allow applications which have data-parallel computations to achieve better performance when compared to traditional CPU systems. However, the GPU is not perfect for all types of computation. The massively parallel SIMT architecture of the GPU can still be constraining in terms of achievable performance. GPU-based systems will typically only be able to achieve between 40%-60% of their peak performance. One of the major problems affecting this effeciency is the GPU memory system, which is tailored to the needs of graphics workloads instead of general-purpose computation.
This thesis intends to show the importance of memory optimizations for GPU systems. In particular, this work addresses problems of data transfer and global atomic memory contention. Using the novel AMD Fusion architecture, we gain overall performance improvements over discrete GPU systems for data-intensive applications. The fused architecture systems offer an interesting trade off by increasing data transfer rates at the cost of some raw computational power. We characterize the performance of different memory paths that are possible because of the shared memory space present on the fused architecture. In addition, we provide a theoretical model which can be used to correctly predict the comparative performance of memory movement techniques for a given data-intensive application and system. In terms of global atomic memory contention, we show improvements in scalability and performance for global synchronization primitives by avoiding contentious global atomic memory accesses. In general, this work shows the importance of understanding the memory system of the GPU architecture to achieve better application performance.
Filename Size Approximate Download Time (Hours:Minutes:Seconds)
28.8 Modem 56K Modem ISDN (64 Kb) ISDN (128 Kb) Higher-speed Access Lee_KS_T_2012.pdf 4.11 Mb 00:19:02 00:09:47 00:08:34 00:04:17 00:00:21
If you have questions or technical problems, please Contact DLA.