Anatomy of a message in the Alewife multiprocessor
Shared-memory provides a uniform and attractive mechanism for communication. For efficiency, it is often implemented with a layer of interpretive hardware on top of a message-passing communications network. This interpretive layer is responsible for data location, data movement, and cache coherence....
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | en_US |
| Published: |
Association for Computing Machinery (ACM)
2015
|
| Online Access: | http://hdl.handle.net/1721.1/99692 https://orcid.org/0000-0002-7015-4262 |
| _version_ | 1826189689877430272 |
|---|---|
| author | Kubiatowicz, John Agarwal, Anant |
| author2 | Massachusetts Institute of Technology. Laboratory for Computer Science |
| author_facet | Massachusetts Institute of Technology. Laboratory for Computer Science Kubiatowicz, John Agarwal, Anant |
| author_sort | Kubiatowicz, John |
| collection | MIT |
| description | Shared-memory provides a uniform and attractive mechanism for communication. For efficiency, it is often implemented with a layer of interpretive hardware on top of a message-passing communications network. This interpretive layer is responsible for data location, data movement, and cache coherence. It uses patterns of communication that benefit common programming styles, but which are only heuristics. This suggests that certain styles of communication may benefit from direct access to the underlying communications substrate. The Alewife machine, a shared-memory multiprocessor being built at MIT, provides such an interface. The interface is an integral part of the shared memory implementation and affords direct, user-level access to the network queues, supports an efficient DMA mechanism, and includes fast trap handling for message reception. This paper discusses the design and implementation of the Alewife message-passing interface and addresses the issues and advantages of using such an interface to complement hardware-synthesized shared memory. |
| first_indexed | 2024-09-23T08:19:34Z |
| format | Article |
| id | mit-1721.1/99692 |
| institution | Massachusetts Institute of Technology |
| language | en_US |
| last_indexed | 2024-09-23T08:19:34Z |
| publishDate | 2015 |
| publisher | Association for Computing Machinery (ACM) |
| record_format | dspace |
| spelling | mit-1721.1/996922022-09-30T09:00:46Z Anatomy of a message in the Alewife multiprocessor Kubiatowicz, John Agarwal, Anant Massachusetts Institute of Technology. Laboratory for Computer Science Kubiatowicz, John Agarwal, Anant Shared-memory provides a uniform and attractive mechanism for communication. For efficiency, it is often implemented with a layer of interpretive hardware on top of a message-passing communications network. This interpretive layer is responsible for data location, data movement, and cache coherence. It uses patterns of communication that benefit common programming styles, but which are only heuristics. This suggests that certain styles of communication may benefit from direct access to the underlying communications substrate. The Alewife machine, a shared-memory multiprocessor being built at MIT, provides such an interface. The interface is an integral part of the shared memory implementation and affords direct, user-level access to the network queues, supports an efficient DMA mechanism, and includes fast trap handling for message reception. This paper discusses the design and implementation of the Alewife message-passing interface and addresses the issues and advantages of using such an interface to complement hardware-synthesized shared memory. National Science Foundation (U.S.) (Grant MIP-9012773) United States. Defense Advanced Research Projects Agency (Contract N00014-87-K-0825) 2015-11-03T19:03:13Z 2015-11-03T19:03:13Z 2014 Article http://purl.org/eprint/type/ConferencePaper 9781450328401 http://hdl.handle.net/1721.1/99692 John Kubiatowicz and Anant Agarwal. 1993. Anatomy of a message in the Alewife multiprocessor. In ACM International Conference on Supercomputing 25th Anniversary Volume, Utpal Banerjee (Ed.). ACM, New York, NY, USA, 193-204. https://orcid.org/0000-0002-7015-4262 en_US http://dx.doi.org/10.1145/2591635.2667168 ACM International Conference on Supercomputing 25th Anniversary Volume Creative Commons Attribution-Noncommercial-Share Alike http://creativecommons.org/licenses/by-nc-sa/4.0/ application/pdf Association for Computing Machinery (ACM) Other univ. web domain |
| spellingShingle | Kubiatowicz, John Agarwal, Anant Anatomy of a message in the Alewife multiprocessor |
| title | Anatomy of a message in the Alewife multiprocessor |
| title_full | Anatomy of a message in the Alewife multiprocessor |
| title_fullStr | Anatomy of a message in the Alewife multiprocessor |
| title_full_unstemmed | Anatomy of a message in the Alewife multiprocessor |
| title_short | Anatomy of a message in the Alewife multiprocessor |
| title_sort | anatomy of a message in the alewife multiprocessor |
| url | http://hdl.handle.net/1721.1/99692 https://orcid.org/0000-0002-7015-4262 |
| work_keys_str_mv | AT kubiatowiczjohn anatomyofamessageinthealewifemultiprocessor AT agarwalanant anatomyofamessageinthealewifemultiprocessor |