Neighborhood-aware data locality optimization for NoC-based multicores

Mahmut Kandemir, Yuanrui Zhang, Jun Liu, Taylan Yemliha

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

7 Citations (Scopus)

Abstract

Data locality optimization is a critical issue for NoC (network-on-chip) based multicore systems. In this paper, focusing on a two-dimensional NoC-based multicore and dataintensive multithreaded applications, we first discuss a data locality aware scheduling algorithm for any given computation-to-core mapping, and then propose an integrated mapping+scheduling algorithm that performs both tasks together. Both our algorithms consider temporal (time-wise) and spatial (neighborhood-aware) data reuse, and try to minimize distance-to-data in on-chip cache accesses. We test the effectiveness of our compiler algorithms using a set of twelve application programs. Our experiments indicate that the proposed algorithms achieve significant improvements in data access latencies (42.7% on average) and overall execution times (24.1% on average). We also conduct a sensitivity analysis where we change the number of cores, on-chip cache capacities, and data movement (migration) strategies. These experiments show that our proposed algorithms generate consistently good results.

Original languageEnglish (US)
Title of host publicationProceedings - International Symposium on Code Generation and Optimization, CGO 2011
Pages191-200
Number of pages10
DOIs
StatePublished - May 30 2011
Event9th International Symposium on Code Generation and Optimization, CGO 2011 - Chamonix, France
Duration: Apr 2 2011Apr 6 2011

Publication series

NameProceedings - International Symposium on Code Generation and Optimization, CGO 2011

Other

Other9th International Symposium on Code Generation and Optimization, CGO 2011
CountryFrance
CityChamonix
Period4/2/114/6/11

Fingerprint

Data Locality
Optimization
Scheduling algorithms
Scheduling Algorithm
Cache
Chip
Data Reuse
Application programs
Compiler
Execution Time
Sensitivity analysis
Migration
Experiment
Sensitivity Analysis
Latency
Experiments
Minimise
Network on chip
Network-on-chip

All Science Journal Classification (ASJC) codes

  • Computational Theory and Mathematics
  • Applied Mathematics

Cite this

Kandemir, M., Zhang, Y., Liu, J., & Yemliha, T. (2011). Neighborhood-aware data locality optimization for NoC-based multicores. In Proceedings - International Symposium on Code Generation and Optimization, CGO 2011 (pp. 191-200). [5764687] (Proceedings - International Symposium on Code Generation and Optimization, CGO 2011). https://doi.org/10.1109/CGO.2011.5764687
Kandemir, Mahmut ; Zhang, Yuanrui ; Liu, Jun ; Yemliha, Taylan. / Neighborhood-aware data locality optimization for NoC-based multicores. Proceedings - International Symposium on Code Generation and Optimization, CGO 2011. 2011. pp. 191-200 (Proceedings - International Symposium on Code Generation and Optimization, CGO 2011).
@inproceedings{37cfb3d61dff42bb9e1b0f0639754a58,
title = "Neighborhood-aware data locality optimization for NoC-based multicores",
abstract = "Data locality optimization is a critical issue for NoC (network-on-chip) based multicore systems. In this paper, focusing on a two-dimensional NoC-based multicore and dataintensive multithreaded applications, we first discuss a data locality aware scheduling algorithm for any given computation-to-core mapping, and then propose an integrated mapping+scheduling algorithm that performs both tasks together. Both our algorithms consider temporal (time-wise) and spatial (neighborhood-aware) data reuse, and try to minimize distance-to-data in on-chip cache accesses. We test the effectiveness of our compiler algorithms using a set of twelve application programs. Our experiments indicate that the proposed algorithms achieve significant improvements in data access latencies (42.7{\%} on average) and overall execution times (24.1{\%} on average). We also conduct a sensitivity analysis where we change the number of cores, on-chip cache capacities, and data movement (migration) strategies. These experiments show that our proposed algorithms generate consistently good results.",
author = "Mahmut Kandemir and Yuanrui Zhang and Jun Liu and Taylan Yemliha",
year = "2011",
month = "5",
day = "30",
doi = "10.1109/CGO.2011.5764687",
language = "English (US)",
isbn = "9781612843551",
series = "Proceedings - International Symposium on Code Generation and Optimization, CGO 2011",
pages = "191--200",
booktitle = "Proceedings - International Symposium on Code Generation and Optimization, CGO 2011",

}

Kandemir, M, Zhang, Y, Liu, J & Yemliha, T 2011, Neighborhood-aware data locality optimization for NoC-based multicores. in Proceedings - International Symposium on Code Generation and Optimization, CGO 2011., 5764687, Proceedings - International Symposium on Code Generation and Optimization, CGO 2011, pp. 191-200, 9th International Symposium on Code Generation and Optimization, CGO 2011, Chamonix, France, 4/2/11. https://doi.org/10.1109/CGO.2011.5764687

Neighborhood-aware data locality optimization for NoC-based multicores. / Kandemir, Mahmut; Zhang, Yuanrui; Liu, Jun; Yemliha, Taylan.

Proceedings - International Symposium on Code Generation and Optimization, CGO 2011. 2011. p. 191-200 5764687 (Proceedings - International Symposium on Code Generation and Optimization, CGO 2011).

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

TY - GEN

T1 - Neighborhood-aware data locality optimization for NoC-based multicores

AU - Kandemir, Mahmut

AU - Zhang, Yuanrui

AU - Liu, Jun

AU - Yemliha, Taylan

PY - 2011/5/30

Y1 - 2011/5/30

N2 - Data locality optimization is a critical issue for NoC (network-on-chip) based multicore systems. In this paper, focusing on a two-dimensional NoC-based multicore and dataintensive multithreaded applications, we first discuss a data locality aware scheduling algorithm for any given computation-to-core mapping, and then propose an integrated mapping+scheduling algorithm that performs both tasks together. Both our algorithms consider temporal (time-wise) and spatial (neighborhood-aware) data reuse, and try to minimize distance-to-data in on-chip cache accesses. We test the effectiveness of our compiler algorithms using a set of twelve application programs. Our experiments indicate that the proposed algorithms achieve significant improvements in data access latencies (42.7% on average) and overall execution times (24.1% on average). We also conduct a sensitivity analysis where we change the number of cores, on-chip cache capacities, and data movement (migration) strategies. These experiments show that our proposed algorithms generate consistently good results.

AB - Data locality optimization is a critical issue for NoC (network-on-chip) based multicore systems. In this paper, focusing on a two-dimensional NoC-based multicore and dataintensive multithreaded applications, we first discuss a data locality aware scheduling algorithm for any given computation-to-core mapping, and then propose an integrated mapping+scheduling algorithm that performs both tasks together. Both our algorithms consider temporal (time-wise) and spatial (neighborhood-aware) data reuse, and try to minimize distance-to-data in on-chip cache accesses. We test the effectiveness of our compiler algorithms using a set of twelve application programs. Our experiments indicate that the proposed algorithms achieve significant improvements in data access latencies (42.7% on average) and overall execution times (24.1% on average). We also conduct a sensitivity analysis where we change the number of cores, on-chip cache capacities, and data movement (migration) strategies. These experiments show that our proposed algorithms generate consistently good results.

UR - http://www.scopus.com/inward/record.url?scp=79957447964&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=79957447964&partnerID=8YFLogxK

U2 - 10.1109/CGO.2011.5764687

DO - 10.1109/CGO.2011.5764687

M3 - Conference contribution

SN - 9781612843551

T3 - Proceedings - International Symposium on Code Generation and Optimization, CGO 2011

SP - 191

EP - 200

BT - Proceedings - International Symposium on Code Generation and Optimization, CGO 2011

ER -

Kandemir M, Zhang Y, Liu J, Yemliha T. Neighborhood-aware data locality optimization for NoC-based multicores. In Proceedings - International Symposium on Code Generation and Optimization, CGO 2011. 2011. p. 191-200. 5764687. (Proceedings - International Symposium on Code Generation and Optimization, CGO 2011). https://doi.org/10.1109/CGO.2011.5764687