Replica: A Wireless Manycore for Communication-Intensive and Approximate Data

Vimuth Fernando, Antonio Franques, Sergi Abadal, Sasa Misailovic, Josep Torrellas

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Data access patterns that involve fine-grained sharing, multicasts, or reductions have proved to be hard to scale in sharedmemory platforms. Recently, wireless on-chip communication has been proposed as a solution to this problem, but a previous architecture has used it only to speed-up synchronization. An intriguing question is whether wireless communication can be widely effective for ordinary shared data. This paper presents Replica, a manycore that uses wireless communication for communication-intensive ordinary data. To deliver high performance, Replica supports an adaptive wireless protocol and selective message dropping. We describe the computational patterns that leverage wireless communication, programming techniques to restructure applications, and tools that help with automation. Our results show that wireless communication is effective for ordinary data. For 64 cores, Replica obtains a mean speed-up of 1.76x over a conventional machine. The mean speed-up reaches 1.89x if approximate-computing transformations are enabled. The average energy consumption is substantially reduced by 34% (or 38% with approximate transformations), and the area increases only modestly.

Original languageEnglish (US)
Title of host publicationASPLOS 2019 - 24th International Conference on Architectural Support for Programming Languages and Operating Systems
PublisherAssociation for Computing Machinery
Pages849-863
Number of pages15
ISBN (Electronic)9781450362405
DOIs
StatePublished - Apr 4 2019
Event24th International Conference on Architectural Support for Programming Languages and Operating Systems, ASPLOS 2019 - Providence, United States
Duration: Apr 13 2019Apr 17 2019

Publication series

NameInternational Conference on Architectural Support for Programming Languages and Operating Systems - ASPLOS

Conference

Conference24th International Conference on Architectural Support for Programming Languages and Operating Systems, ASPLOS 2019
CountryUnited States
CityProvidence
Period4/13/194/17/19

Fingerprint

Communication
Synchronization
Energy utilization
Automation

Keywords

  • Approximate
  • Multicore
  • Parallelism
  • Wireless

ASJC Scopus subject areas

  • Software
  • Information Systems
  • Hardware and Architecture

Cite this

Fernando, V., Franques, A., Abadal, S., Misailovic, S., & Torrellas, J. (2019). Replica: A Wireless Manycore for Communication-Intensive and Approximate Data. In ASPLOS 2019 - 24th International Conference on Architectural Support for Programming Languages and Operating Systems (pp. 849-863). (International Conference on Architectural Support for Programming Languages and Operating Systems - ASPLOS). Association for Computing Machinery. https://doi.org/10.1145/3297858.3304033

Replica : A Wireless Manycore for Communication-Intensive and Approximate Data. / Fernando, Vimuth; Franques, Antonio; Abadal, Sergi; Misailovic, Sasa; Torrellas, Josep.

ASPLOS 2019 - 24th International Conference on Architectural Support for Programming Languages and Operating Systems. Association for Computing Machinery, 2019. p. 849-863 (International Conference on Architectural Support for Programming Languages and Operating Systems - ASPLOS).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Fernando, V, Franques, A, Abadal, S, Misailovic, S & Torrellas, J 2019, Replica: A Wireless Manycore for Communication-Intensive and Approximate Data. in ASPLOS 2019 - 24th International Conference on Architectural Support for Programming Languages and Operating Systems. International Conference on Architectural Support for Programming Languages and Operating Systems - ASPLOS, Association for Computing Machinery, pp. 849-863, 24th International Conference on Architectural Support for Programming Languages and Operating Systems, ASPLOS 2019, Providence, United States, 4/13/19. https://doi.org/10.1145/3297858.3304033
Fernando V, Franques A, Abadal S, Misailovic S, Torrellas J. Replica: A Wireless Manycore for Communication-Intensive and Approximate Data. In ASPLOS 2019 - 24th International Conference on Architectural Support for Programming Languages and Operating Systems. Association for Computing Machinery. 2019. p. 849-863. (International Conference on Architectural Support for Programming Languages and Operating Systems - ASPLOS). https://doi.org/10.1145/3297858.3304033
Fernando, Vimuth ; Franques, Antonio ; Abadal, Sergi ; Misailovic, Sasa ; Torrellas, Josep. / Replica : A Wireless Manycore for Communication-Intensive and Approximate Data. ASPLOS 2019 - 24th International Conference on Architectural Support for Programming Languages and Operating Systems. Association for Computing Machinery, 2019. pp. 849-863 (International Conference on Architectural Support for Programming Languages and Operating Systems - ASPLOS).
@inproceedings{0ad6b336b8bf4ca0b1854eac1f5a374c,
title = "Replica: A Wireless Manycore for Communication-Intensive and Approximate Data",
abstract = "Data access patterns that involve fine-grained sharing, multicasts, or reductions have proved to be hard to scale in sharedmemory platforms. Recently, wireless on-chip communication has been proposed as a solution to this problem, but a previous architecture has used it only to speed-up synchronization. An intriguing question is whether wireless communication can be widely effective for ordinary shared data. This paper presents Replica, a manycore that uses wireless communication for communication-intensive ordinary data. To deliver high performance, Replica supports an adaptive wireless protocol and selective message dropping. We describe the computational patterns that leverage wireless communication, programming techniques to restructure applications, and tools that help with automation. Our results show that wireless communication is effective for ordinary data. For 64 cores, Replica obtains a mean speed-up of 1.76x over a conventional machine. The mean speed-up reaches 1.89x if approximate-computing transformations are enabled. The average energy consumption is substantially reduced by 34{\%} (or 38{\%} with approximate transformations), and the area increases only modestly.",
keywords = "Approximate, Multicore, Parallelism, Wireless",
author = "Vimuth Fernando and Antonio Franques and Sergi Abadal and Sasa Misailovic and Josep Torrellas",
year = "2019",
month = "4",
day = "4",
doi = "10.1145/3297858.3304033",
language = "English (US)",
series = "International Conference on Architectural Support for Programming Languages and Operating Systems - ASPLOS",
publisher = "Association for Computing Machinery",
pages = "849--863",
booktitle = "ASPLOS 2019 - 24th International Conference on Architectural Support for Programming Languages and Operating Systems",

}

TY - GEN

T1 - Replica

T2 - A Wireless Manycore for Communication-Intensive and Approximate Data

AU - Fernando, Vimuth

AU - Franques, Antonio

AU - Abadal, Sergi

AU - Misailovic, Sasa

AU - Torrellas, Josep

PY - 2019/4/4

Y1 - 2019/4/4

N2 - Data access patterns that involve fine-grained sharing, multicasts, or reductions have proved to be hard to scale in sharedmemory platforms. Recently, wireless on-chip communication has been proposed as a solution to this problem, but a previous architecture has used it only to speed-up synchronization. An intriguing question is whether wireless communication can be widely effective for ordinary shared data. This paper presents Replica, a manycore that uses wireless communication for communication-intensive ordinary data. To deliver high performance, Replica supports an adaptive wireless protocol and selective message dropping. We describe the computational patterns that leverage wireless communication, programming techniques to restructure applications, and tools that help with automation. Our results show that wireless communication is effective for ordinary data. For 64 cores, Replica obtains a mean speed-up of 1.76x over a conventional machine. The mean speed-up reaches 1.89x if approximate-computing transformations are enabled. The average energy consumption is substantially reduced by 34% (or 38% with approximate transformations), and the area increases only modestly.

AB - Data access patterns that involve fine-grained sharing, multicasts, or reductions have proved to be hard to scale in sharedmemory platforms. Recently, wireless on-chip communication has been proposed as a solution to this problem, but a previous architecture has used it only to speed-up synchronization. An intriguing question is whether wireless communication can be widely effective for ordinary shared data. This paper presents Replica, a manycore that uses wireless communication for communication-intensive ordinary data. To deliver high performance, Replica supports an adaptive wireless protocol and selective message dropping. We describe the computational patterns that leverage wireless communication, programming techniques to restructure applications, and tools that help with automation. Our results show that wireless communication is effective for ordinary data. For 64 cores, Replica obtains a mean speed-up of 1.76x over a conventional machine. The mean speed-up reaches 1.89x if approximate-computing transformations are enabled. The average energy consumption is substantially reduced by 34% (or 38% with approximate transformations), and the area increases only modestly.

KW - Approximate

KW - Multicore

KW - Parallelism

KW - Wireless

UR - http://www.scopus.com/inward/record.url?scp=85064622316&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85064622316&partnerID=8YFLogxK

U2 - 10.1145/3297858.3304033

DO - 10.1145/3297858.3304033

M3 - Conference contribution

AN - SCOPUS:85064622316

T3 - International Conference on Architectural Support for Programming Languages and Operating Systems - ASPLOS

SP - 849

EP - 863

BT - ASPLOS 2019 - 24th International Conference on Architectural Support for Programming Languages and Operating Systems

PB - Association for Computing Machinery

ER -