The minos computing library: Efficient parallel programming for extremely heterogeneous systems

Roberto Gioiosa; Burcu O. Mutlu; Seyong Lee; Jeffrey S. Vetter; Giulio Picierro; Marco Cesati

doi:10.1145/3366428.3380770

The minos computing library: Efficient parallel programming for extremely heterogeneous systems

Roberto Gioiosa, Burcu O. Mutlu, Seyong Lee, Jeffrey S. Vetter, Giulio Picierro, Marco Cesati

Architectures and Performance

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

12 Scopus citations

Abstract

Hardware specialization has become the silver bullet to achieve efficient high performance, from Systems-on-Chip systems, where hardware specialization can be "extreme", to large-scale HPC systems. As the complexity of the systems increases, so does the complexity of programming such architectures in a portable way. This work introduces the Minos Computing Library (MCL), as system software, programming model, and programming model runtime that facilitate programming extremely heterogeneous systems. MCL supports the execution of several multi-threaded applications within the same compute node, performs asynchronous execution of application tasks, efficiently balances computation across hardware resources, and provides performance portability. We show that code developed on a personal desktop automatically scales up to fully utilize powerful workstations with 8 GPUs and down to power-efficient embedded systems. MCL provides up to 17.5x speedup over OpenCL on NVIDIA DGX-1 systems and up to 1.88x speedup on single-GPU systems. In multi-application workloads, MCL's dynamic resource allocation provides up to 2.43x performance improvement over manual, static resources allocation.

Original language	English
Title of host publication	GPGPU 2020 - Proceedings of the 2020 General Purpose Processing Using GPU
Publisher	Association for Computing Machinery, Inc
Pages	1-10
Number of pages	10
ISBN (Electronic)	9781450370257
DOIs	https://doi.org/10.1145/3366428.3380770
State	Published - Feb 23 2020
Event	13th Annual Workshop on General Purpose Processing using Graphics Processing Unit, GPGPU 2020 - San Diego, United States Duration: Feb 23 2020 → Feb 23 2020

Publication series

Name	GPGPU 2020 - Proceedings of the 2020 General Purpose Processing Using GPU

Conference

Conference	13th Annual Workshop on General Purpose Processing using Graphics Processing Unit, GPGPU 2020
Country/Territory	United States
City	San Diego
Period	02/23/20 → 02/23/20

Funding

This research is supported by the Department of Energy (DOE) Advanced Scientific Computing Research (ASCR) LAB 19-2119 and Defense Advanced Research Projects Agency (DARPA) HR001117S0055, Program Area Domain-Specific System on Chip (DSSoC).

Keywords

Asynchronous runtime
GPU
Heterogeneous systems
System software
Task-based runtime

Access to Document

10.1145/3366428.3380770

Cite this

Gioiosa, R., Mutlu, B. O., Lee, S., Vetter, J. S., Picierro, G., & Cesati, M. (2020). The minos computing library: Efficient parallel programming for extremely heterogeneous systems. In GPGPU 2020 - Proceedings of the 2020 General Purpose Processing Using GPU (pp. 1-10). (GPGPU 2020 - Proceedings of the 2020 General Purpose Processing Using GPU). Association for Computing Machinery, Inc. https://doi.org/10.1145/3366428.3380770

@inproceedings{2b9a463d76614f0084d077e8d196bddc,

title = "The minos computing library: Efficient parallel programming for extremely heterogeneous systems",

abstract = "Hardware specialization has become the silver bullet to achieve efficient high performance, from Systems-on-Chip systems, where hardware specialization can be {"}extreme{"}, to large-scale HPC systems. As the complexity of the systems increases, so does the complexity of programming such architectures in a portable way. This work introduces the Minos Computing Library (MCL), as system software, programming model, and programming model runtime that facilitate programming extremely heterogeneous systems. MCL supports the execution of several multi-threaded applications within the same compute node, performs asynchronous execution of application tasks, efficiently balances computation across hardware resources, and provides performance portability. We show that code developed on a personal desktop automatically scales up to fully utilize powerful workstations with 8 GPUs and down to power-efficient embedded systems. MCL provides up to 17.5x speedup over OpenCL on NVIDIA DGX-1 systems and up to 1.88x speedup on single-GPU systems. In multi-application workloads, MCL's dynamic resource allocation provides up to 2.43x performance improvement over manual, static resources allocation.",

keywords = "Asynchronous runtime, GPU, Heterogeneous systems, System software, Task-based runtime",

author = "Roberto Gioiosa and Mutlu, {Burcu O.} and Seyong Lee and Vetter, {Jeffrey S.} and Giulio Picierro and Marco Cesati",

note = "Publisher Copyright: {\textcopyright} 2020 Association for Computing Machinery.; 13th Annual Workshop on General Purpose Processing using Graphics Processing Unit, GPGPU 2020 ; Conference date: 23-02-2020 Through 23-02-2020",

year = "2020",

month = feb,

day = "23",

doi = "10.1145/3366428.3380770",

language = "English",

series = "GPGPU 2020 - Proceedings of the 2020 General Purpose Processing Using GPU",

publisher = "Association for Computing Machinery, Inc",

pages = "1--10",

booktitle = "GPGPU 2020 - Proceedings of the 2020 General Purpose Processing Using GPU",

}

Gioiosa, R, Mutlu, BO, Lee, S , Vetter, JS, Picierro, G & Cesati, M 2020, The minos computing library: Efficient parallel programming for extremely heterogeneous systems. in GPGPU 2020 - Proceedings of the 2020 General Purpose Processing Using GPU. GPGPU 2020 - Proceedings of the 2020 General Purpose Processing Using GPU, Association for Computing Machinery, Inc, pp. 1-10, 13th Annual Workshop on General Purpose Processing using Graphics Processing Unit, GPGPU 2020, San Diego, United States, 02/23/20. https://doi.org/10.1145/3366428.3380770

The minos computing library: Efficient parallel programming for extremely heterogeneous systems. / Gioiosa, Roberto; Mutlu, Burcu O.; Lee, Seyong et al.
GPGPU 2020 - Proceedings of the 2020 General Purpose Processing Using GPU. Association for Computing Machinery, Inc, 2020. p. 1-10 (GPGPU 2020 - Proceedings of the 2020 General Purpose Processing Using GPU).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - The minos computing library

T2 - 13th Annual Workshop on General Purpose Processing using Graphics Processing Unit, GPGPU 2020

AU - Gioiosa, Roberto

AU - Mutlu, Burcu O.

AU - Lee, Seyong

AU - Vetter, Jeffrey S.

AU - Picierro, Giulio

AU - Cesati, Marco

PY - 2020/2/23

Y1 - 2020/2/23

N2 - Hardware specialization has become the silver bullet to achieve efficient high performance, from Systems-on-Chip systems, where hardware specialization can be "extreme", to large-scale HPC systems. As the complexity of the systems increases, so does the complexity of programming such architectures in a portable way. This work introduces the Minos Computing Library (MCL), as system software, programming model, and programming model runtime that facilitate programming extremely heterogeneous systems. MCL supports the execution of several multi-threaded applications within the same compute node, performs asynchronous execution of application tasks, efficiently balances computation across hardware resources, and provides performance portability. We show that code developed on a personal desktop automatically scales up to fully utilize powerful workstations with 8 GPUs and down to power-efficient embedded systems. MCL provides up to 17.5x speedup over OpenCL on NVIDIA DGX-1 systems and up to 1.88x speedup on single-GPU systems. In multi-application workloads, MCL's dynamic resource allocation provides up to 2.43x performance improvement over manual, static resources allocation.

AB - Hardware specialization has become the silver bullet to achieve efficient high performance, from Systems-on-Chip systems, where hardware specialization can be "extreme", to large-scale HPC systems. As the complexity of the systems increases, so does the complexity of programming such architectures in a portable way. This work introduces the Minos Computing Library (MCL), as system software, programming model, and programming model runtime that facilitate programming extremely heterogeneous systems. MCL supports the execution of several multi-threaded applications within the same compute node, performs asynchronous execution of application tasks, efficiently balances computation across hardware resources, and provides performance portability. We show that code developed on a personal desktop automatically scales up to fully utilize powerful workstations with 8 GPUs and down to power-efficient embedded systems. MCL provides up to 17.5x speedup over OpenCL on NVIDIA DGX-1 systems and up to 1.88x speedup on single-GPU systems. In multi-application workloads, MCL's dynamic resource allocation provides up to 2.43x performance improvement over manual, static resources allocation.

KW - Asynchronous runtime

KW - GPU

KW - Heterogeneous systems

KW - System software

KW - Task-based runtime

UR - http://www.scopus.com/inward/record.url?scp=85082054349&partnerID=8YFLogxK

U2 - 10.1145/3366428.3380770

DO - 10.1145/3366428.3380770

M3 - Conference contribution

AN - SCOPUS:85082054349

T3 - GPGPU 2020 - Proceedings of the 2020 General Purpose Processing Using GPU

SP - 1

EP - 10

BT - GPGPU 2020 - Proceedings of the 2020 General Purpose Processing Using GPU

PB - Association for Computing Machinery, Inc

Y2 - 23 February 2020 through 23 February 2020

ER -

Gioiosa R, Mutlu BO, Lee S , Vetter JS, Picierro G, Cesati M. The minos computing library: Efficient parallel programming for extremely heterogeneous systems. In GPGPU 2020 - Proceedings of the 2020 General Purpose Processing Using GPU. Association for Computing Machinery, Inc. 2020. p. 1-10. (GPGPU 2020 - Proceedings of the 2020 General Purpose Processing Using GPU). doi: 10.1145/3366428.3380770

The minos computing library: Efficient parallel programming for extremely heterogeneous systems

Abstract

Publication series

Conference

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this