The Multi-Threaded Architecture Multiprocessor (1994)
BibTeX
@MISC{Hum94themulti-threaded,
author = {Herbert Hum and Herbert H. J and Humy Olivier and Olivier Maquelin and Kevin Theobald and Xinan Tang and Xinan Tang and Guang Gao and Phil Cupryk and Nasser Elmasri and Laurie Hendren and Alberto Jimenez and Shoba Krishnany and Andres Marquez and Shamir Merali and Shashank Nemawarkar and Prakash Panangaden and Xun Xue and Yingchun Zhu},
title = {The Multi-Threaded Architecture Multiprocessor},
year = {1994}
}
Bookmark
 |
 |
 |
 |
 |
 |
OpenURL
Abstract
Multithreaded node architectures have been proposed for future multiprocessor systems. However, some open issues remain: can efficient multithreading support be provided in a multiprocessor machine such that it is capable of tolerating the synchronization and communication latencies, without intruding on the performance of sequentially-executed code? And how much (quantitatively) does such non-intrusive multithreading support contribute to the scalable parallel performance in the presence of increasing interprocessor communication and synchronization demands? In this paper, we describe the design of the MTA (Multi-Threaded Architecture) model which attempts to address the above issues. We have implemented the MTA model on an experimental high-performance multiprocessor testbed called MANNA 2.0. The MTA-MANNA platform has for each processing node, an Execution Unit (EU) for performing the computations and is programmed as a standard superscalar processor, and a Synchronization Unit (SU)...
