The MIT Alewife Machine: A Large-Scale Distributed-Memory Multiprocessor (1991)
| Venue: | In Proceedings of Workshop on Scalable Shared Memory Multiprocessors |
| Citations: | 152 - 25 self |
BibTeX
@INPROCEEDINGS{Agarwal91themit,
author = {Anant Agarwal and David Chaiken and David Kranz and John Kubiatowicz and Kiyoshi Kurihara and Gino Maa and Dan Nussbaum and Mike Parkin and Donald Yeung},
title = {The MIT Alewife Machine: A Large-Scale Distributed-Memory Multiprocessor},
booktitle = {In Proceedings of Workshop on Scalable Shared Memory Multiprocessors},
year = {1991},
publisher = {Kluwer Academic Publishers}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
The Alewife multiprocessor project focuses on the architecture and design of a large-scale parallel machine. The machine uses a low-dimensional direct interconnection network to provide scalable communication bandwidth, while allowing the exploitation of locality. Despite its distributed-memory architecture, Alewife allows efficient shared-memory programming through a multilayered approach to locality management. A new scalable cache-coherence scheme called LimitLESS directories allows the use of caches for reducing communication latency and network bandwidth requirements. Alewife also employs run-time and compile-time methods for partitioning and placement of data and processes to enhance communication locality. While the above methods attempt to minimize communication latency, communication with distant processors cannot be completely avoided. Alewife's processor, Sparcle, is designed to tolerate these latencies by rapidly switching between threads of computation. This paper describe...
Keyphrases
large-scale distributed-memory multiprocessor mit alewife machine communication latency large-scale parallel machine paper describe new scalable cache-coherence scheme multilayered approach alewife multiprocessor project communication locality locality management efficient shared-memory programming distant processor distributed-memory architecture compile-time method low-dimensional direct interconnection network limitless directory network bandwidth requirement scalable communication bandwidth method attempt
