A quantitative comparison of PRAM based emulated shared memory architectures to current multicore CPUs and GPUs

Erik Hansson; Erik Alnervik; Christoph Kessler; Martti Forsell

A quantitative comparison of PRAM based emulated shared memory architectures to current multicore CPUs and GPUs

Erik Hansson, Erik Alnervik, Christoph Kessler, Martti Forsell

University of Linköping

Research output: Chapter in Book/Report/Conference proceeding › Conference article in proceedings › Scientific › peer-review

Abstract

The performance of current multicore CPUs and GPUs is limited in computations making frequent use of com- munication/synchronization between the subtasks executed in parallel. This is because the directory-based cache systems scale weakly and/or the cost of synchronization is high. The Emulated Shared Memory (ESM) architectures relying on multithreading and efficient synchronization mechanism have been developed to solve these problems affecting both performance and pro- grammability of current machines. In this paper, we compare preliminarily the performance of three hardware implemented ESM architectures with state-of-the-art multicore CPUs and GPUs. The benchmarks are selected to cover different patterns of parallel computation and therefore reveal the performance potential of ESM architectures with respect to current multicores.

Original language	English
Title of host publication	2014 Workshop Proceedings on Architecture of Computing Systems (ARCS 2014)
Publisher	VDE Verlag
Number of pages	7
ISBN (Print)	978-3-8007-3579-2
Publication status	Published - 2014
MoE publication type	A4 Article in a conference publication
Event	27th International Conference on Architecture of Computing Systems, ARCS 2014 - Lübeck, Germany Duration: 25 Feb 2014 → 28 Feb 2014

Conference

Conference	27th International Conference on Architecture of Computing Systems, ARCS 2014
Abbreviated title	ARCS 2014
Country/Territory	Germany
City	Lübeck
Period	25/02/14 → 28/02/14

Access to Document

https://ieeexplore.ieee.org/document/6775082

Cite this

@inproceedings{864b3c4f6c46468bb13cf63a1bbbbb1d,

title = "A quantitative comparison of PRAM based emulated shared memory architectures to current multicore CPUs and GPUs",

abstract = "The performance of current multicore CPUs and GPUs is limited in computations making frequent use of com- munication/synchronization between the subtasks executed in parallel. This is because the directory-based cache systems scale weakly and/or the cost of synchronization is high. The Emulated Shared Memory (ESM) architectures relying on multithreading and efficient synchronization mechanism have been developed to solve these problems affecting both performance and pro- grammability of current machines. In this paper, we compare preliminarily the performance of three hardware implemented ESM architectures with state-of-the-art multicore CPUs and GPUs. The benchmarks are selected to cover different patterns of parallel computation and therefore reveal the performance potential of ESM architectures with respect to current multicores.",

author = "Erik Hansson and Erik Alnervik and Christoph Kessler and Martti Forsell",

year = "2014",

language = "English",

isbn = "978-3-8007-3579-2",

booktitle = "2014 Workshop Proceedings on Architecture of Computing Systems (ARCS 2014)",

publisher = "VDE Verlag",

address = "Germany",

note = "27th International Conference on Architecture of Computing Systems, ARCS 2014, ARCS 2014 ; Conference date: 25-02-2014 Through 28-02-2014",

}

Hansson, E, Alnervik, E, Kessler, C & Forsell, M 2014, A quantitative comparison of PRAM based emulated shared memory architectures to current multicore CPUs and GPUs. in 2014 Workshop Proceedings on Architecture of Computing Systems (ARCS 2014). VDE Verlag, 27th International Conference on Architecture of Computing Systems, ARCS 2014, Lübeck, Germany, 25/02/14. <https://ieeexplore.ieee.org/document/6775082>

A quantitative comparison of PRAM based emulated shared memory architectures to current multicore CPUs and GPUs. / Hansson, Erik; Alnervik, Erik; Kessler, Christoph et al.
2014 Workshop Proceedings on Architecture of Computing Systems (ARCS 2014). VDE Verlag, 2014.

Research output: Chapter in Book/Report/Conference proceeding › Conference article in proceedings › Scientific › peer-review

TY - GEN

T1 - A quantitative comparison of PRAM based emulated shared memory architectures to current multicore CPUs and GPUs

AU - Hansson, Erik

AU - Alnervik, Erik

AU - Kessler, Christoph

AU - Forsell, Martti

PY - 2014

Y1 - 2014

N2 - The performance of current multicore CPUs and GPUs is limited in computations making frequent use of com- munication/synchronization between the subtasks executed in parallel. This is because the directory-based cache systems scale weakly and/or the cost of synchronization is high. The Emulated Shared Memory (ESM) architectures relying on multithreading and efficient synchronization mechanism have been developed to solve these problems affecting both performance and pro- grammability of current machines. In this paper, we compare preliminarily the performance of three hardware implemented ESM architectures with state-of-the-art multicore CPUs and GPUs. The benchmarks are selected to cover different patterns of parallel computation and therefore reveal the performance potential of ESM architectures with respect to current multicores.

AB - The performance of current multicore CPUs and GPUs is limited in computations making frequent use of com- munication/synchronization between the subtasks executed in parallel. This is because the directory-based cache systems scale weakly and/or the cost of synchronization is high. The Emulated Shared Memory (ESM) architectures relying on multithreading and efficient synchronization mechanism have been developed to solve these problems affecting both performance and pro- grammability of current machines. In this paper, we compare preliminarily the performance of three hardware implemented ESM architectures with state-of-the-art multicore CPUs and GPUs. The benchmarks are selected to cover different patterns of parallel computation and therefore reveal the performance potential of ESM architectures with respect to current multicores.

M3 - Conference article in proceedings

SN - 978-3-8007-3579-2

BT - 2014 Workshop Proceedings on Architecture of Computing Systems (ARCS 2014)

PB - VDE Verlag

T2 - 27th International Conference on Architecture of Computing Systems, ARCS 2014

Y2 - 25 February 2014 through 28 February 2014

ER -

A quantitative comparison of PRAM based emulated shared memory architectures to current multicore CPUs and GPUs

Abstract

Conference

Access to Document

Fingerprint

Cite this