Collaborative computing for heterogeneous integrated systems

Li Wen Chang, Juan Gómez-Luna, Izzat El Hajj, Sitao Huang, Deming Chen, Wen-Mei W Hwu

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

Abstract

Computing systems today typically employ, in addition to powerful CPUs, various types of specialized devices such as Graphics Processing Units (GPUs) and Field-Programmable Gate Arrays (FPGAs). Such heterogeneous systems are evolving towards tighter integration of CPUs and devices for improved performance and reduced energy consumption. Compared to traditional use of GPUs and FPGAs as offload accelerators, this tight integration enables close collaboration between processors and devices, which is important for better utilization of system resources and higher performance. Programming interfaces are also adapting rapidly to these tightly integrated heterogeneous platforms by introducing features such as shared virtual memory, memory coherence, and system-wide atomics, making collaborative computing even more practical. In this paper, we survey current integrated heterogeneous systems and corresponding collaboration techniques. We evaluate the impact of collaborative computing on two heterogeneous integrated systems, CPU-GPU and CPU-FPGA, using OpenCL. Finally, we discuss the limitation of OpenCL and envision what suitable programming languages for collaborative computing will look like.

Original languageEnglish (US)
Title of host publicationICPE 2017 - Proceedings of the 2017 ACM/SPEC International Conference on Performance Engineering
PublisherAssociation for Computing Machinery, Inc
Pages385-388
Number of pages4
ISBN (Electronic)9781450344043
DOIs
StatePublished - Apr 17 2017
Event8th ACM/SPEC International Conference on Performance Engineering, ICPE 2017 - L'Aquila, Italy
Duration: Apr 22 2017Apr 26 2017

Publication series

NameICPE 2017 - Proceedings of the 2017 ACM/SPEC International Conference on Performance Engineering

Other

Other8th ACM/SPEC International Conference on Performance Engineering, ICPE 2017
CountryItaly
CityL'Aquila
Period4/22/174/26/17

Fingerprint

Computer supported cooperative work
Program processors
Field programmable gate arrays (FPGA)
Computer systems
Data storage equipment
Computer programming languages
Particle accelerators
Energy utilization
Graphics processing unit

ASJC Scopus subject areas

  • Hardware and Architecture
  • Software
  • Computer Science Applications

Cite this

Chang, L. W., Gómez-Luna, J., El Hajj, I., Huang, S., Chen, D., & Hwu, W-M. W. (2017). Collaborative computing for heterogeneous integrated systems. In ICPE 2017 - Proceedings of the 2017 ACM/SPEC International Conference on Performance Engineering (pp. 385-388). (ICPE 2017 - Proceedings of the 2017 ACM/SPEC International Conference on Performance Engineering). Association for Computing Machinery, Inc. https://doi.org/10.1145/3030207.3030244

Collaborative computing for heterogeneous integrated systems. / Chang, Li Wen; Gómez-Luna, Juan; El Hajj, Izzat; Huang, Sitao; Chen, Deming; Hwu, Wen-Mei W.

ICPE 2017 - Proceedings of the 2017 ACM/SPEC International Conference on Performance Engineering. Association for Computing Machinery, Inc, 2017. p. 385-388 (ICPE 2017 - Proceedings of the 2017 ACM/SPEC International Conference on Performance Engineering).

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

Chang, LW, Gómez-Luna, J, El Hajj, I, Huang, S, Chen, D & Hwu, W-MW 2017, Collaborative computing for heterogeneous integrated systems. in ICPE 2017 - Proceedings of the 2017 ACM/SPEC International Conference on Performance Engineering. ICPE 2017 - Proceedings of the 2017 ACM/SPEC International Conference on Performance Engineering, Association for Computing Machinery, Inc, pp. 385-388, 8th ACM/SPEC International Conference on Performance Engineering, ICPE 2017, L'Aquila, Italy, 4/22/17. https://doi.org/10.1145/3030207.3030244
Chang LW, Gómez-Luna J, El Hajj I, Huang S, Chen D, Hwu W-MW. Collaborative computing for heterogeneous integrated systems. In ICPE 2017 - Proceedings of the 2017 ACM/SPEC International Conference on Performance Engineering. Association for Computing Machinery, Inc. 2017. p. 385-388. (ICPE 2017 - Proceedings of the 2017 ACM/SPEC International Conference on Performance Engineering). https://doi.org/10.1145/3030207.3030244
Chang, Li Wen ; Gómez-Luna, Juan ; El Hajj, Izzat ; Huang, Sitao ; Chen, Deming ; Hwu, Wen-Mei W. / Collaborative computing for heterogeneous integrated systems. ICPE 2017 - Proceedings of the 2017 ACM/SPEC International Conference on Performance Engineering. Association for Computing Machinery, Inc, 2017. pp. 385-388 (ICPE 2017 - Proceedings of the 2017 ACM/SPEC International Conference on Performance Engineering).
@inproceedings{c220fe1c06f44d26b3248db95a370937,
title = "Collaborative computing for heterogeneous integrated systems",
abstract = "Computing systems today typically employ, in addition to powerful CPUs, various types of specialized devices such as Graphics Processing Units (GPUs) and Field-Programmable Gate Arrays (FPGAs). Such heterogeneous systems are evolving towards tighter integration of CPUs and devices for improved performance and reduced energy consumption. Compared to traditional use of GPUs and FPGAs as offload accelerators, this tight integration enables close collaboration between processors and devices, which is important for better utilization of system resources and higher performance. Programming interfaces are also adapting rapidly to these tightly integrated heterogeneous platforms by introducing features such as shared virtual memory, memory coherence, and system-wide atomics, making collaborative computing even more practical. In this paper, we survey current integrated heterogeneous systems and corresponding collaboration techniques. We evaluate the impact of collaborative computing on two heterogeneous integrated systems, CPU-GPU and CPU-FPGA, using OpenCL. Finally, we discuss the limitation of OpenCL and envision what suitable programming languages for collaborative computing will look like.",
author = "Chang, {Li Wen} and Juan G{\'o}mez-Luna and {El Hajj}, Izzat and Sitao Huang and Deming Chen and Hwu, {Wen-Mei W}",
year = "2017",
month = "4",
day = "17",
doi = "10.1145/3030207.3030244",
language = "English (US)",
series = "ICPE 2017 - Proceedings of the 2017 ACM/SPEC International Conference on Performance Engineering",
publisher = "Association for Computing Machinery, Inc",
pages = "385--388",
booktitle = "ICPE 2017 - Proceedings of the 2017 ACM/SPEC International Conference on Performance Engineering",

}

TY - GEN

T1 - Collaborative computing for heterogeneous integrated systems

AU - Chang, Li Wen

AU - Gómez-Luna, Juan

AU - El Hajj, Izzat

AU - Huang, Sitao

AU - Chen, Deming

AU - Hwu, Wen-Mei W

PY - 2017/4/17

Y1 - 2017/4/17

N2 - Computing systems today typically employ, in addition to powerful CPUs, various types of specialized devices such as Graphics Processing Units (GPUs) and Field-Programmable Gate Arrays (FPGAs). Such heterogeneous systems are evolving towards tighter integration of CPUs and devices for improved performance and reduced energy consumption. Compared to traditional use of GPUs and FPGAs as offload accelerators, this tight integration enables close collaboration between processors and devices, which is important for better utilization of system resources and higher performance. Programming interfaces are also adapting rapidly to these tightly integrated heterogeneous platforms by introducing features such as shared virtual memory, memory coherence, and system-wide atomics, making collaborative computing even more practical. In this paper, we survey current integrated heterogeneous systems and corresponding collaboration techniques. We evaluate the impact of collaborative computing on two heterogeneous integrated systems, CPU-GPU and CPU-FPGA, using OpenCL. Finally, we discuss the limitation of OpenCL and envision what suitable programming languages for collaborative computing will look like.

AB - Computing systems today typically employ, in addition to powerful CPUs, various types of specialized devices such as Graphics Processing Units (GPUs) and Field-Programmable Gate Arrays (FPGAs). Such heterogeneous systems are evolving towards tighter integration of CPUs and devices for improved performance and reduced energy consumption. Compared to traditional use of GPUs and FPGAs as offload accelerators, this tight integration enables close collaboration between processors and devices, which is important for better utilization of system resources and higher performance. Programming interfaces are also adapting rapidly to these tightly integrated heterogeneous platforms by introducing features such as shared virtual memory, memory coherence, and system-wide atomics, making collaborative computing even more practical. In this paper, we survey current integrated heterogeneous systems and corresponding collaboration techniques. We evaluate the impact of collaborative computing on two heterogeneous integrated systems, CPU-GPU and CPU-FPGA, using OpenCL. Finally, we discuss the limitation of OpenCL and envision what suitable programming languages for collaborative computing will look like.

UR - http://www.scopus.com/inward/record.url?scp=85019016184&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85019016184&partnerID=8YFLogxK

U2 - 10.1145/3030207.3030244

DO - 10.1145/3030207.3030244

M3 - Conference contribution

T3 - ICPE 2017 - Proceedings of the 2017 ACM/SPEC International Conference on Performance Engineering

SP - 385

EP - 388

BT - ICPE 2017 - Proceedings of the 2017 ACM/SPEC International Conference on Performance Engineering

PB - Association for Computing Machinery, Inc

ER -