deposit_your_work

The impact of instruction-level parallelism on multiprocessor performance and simulation methodology

Files in this item

Files Size Format View
1384393.PDF 2.330Mb application/pdf Thumbnail

Show full item record

Item Metadata

Title: The impact of instruction-level parallelism on multiprocessor performance and simulation methodology
Author: Pai, Vijay Sadananda
Advisor: Adve, Sarita V.
Degree: Master of Science thesis
Abstract: Current microprocessors exploit high levels of instruction-level parallelism (ILP). This thesis presents the first detailed analysis of the impact of such processors on shared-memory multiprocessors. We find that ILP techniques substantially reduce CPU time in multiprocessors, but are less effective in reducing memory stall time for our applications. Consequently, despite the latency-tolerating techniques incorporated in ILP processors, memory stall time becomes a larger component of execution time and parallel efficiencies are generally poorer in our ILP-based multiprocessor than in an otherwise equivalent previous-generation multiprocessor. We identify clustering independent read misses together in the processor instruction window as a key optimization to exploit the ILP features of current processors. We also use the above analysis to examine the validity of direct-execution simulators with previous-generation processor models to approximate ILP-based multiprocessors. We find that, with appropriate approximations, such simulators can reasonably characterize the behavior of applications with poor overlap of read misses. However, they can be highly inaccurate for applications with high overlap of read misses.
Citation: Pai, Vijay Sadananda. (1997) "The impact of instruction-level parallelism on multiprocessor performance and simulation methodology." Masters Thesis, Rice University. http://hdl.handle.net/1911/17116.
URI: http://hdl.handle.net/1911/17116
Date: 1997

This item appears in the following Collection(s)