Architecting an LTE base station with graphics processing units

Q. Zheng, Y. Chen, R. Dreslinski, Chaitali Chakrabarti, A. Anastasopoulos, S. Mahlke, T. Mudge

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

8 Citations (Scopus)

Abstract

Due to the rapid growth of mobile communication, wireless base stations are becoming a significant consumer of computational resources. Historically, base stations have been built from ASICs, DSP processors, or FPGAs. This paper studies the feasibility of building wireless base stations from commercial graphics processing units (GPUs). GPUs are attractive because they are widely used massively parallel devices that can be programmed in a high level language. Base station workloads are highly parallel, making GPUs a potential candidate for a cost effective programmable solution. In this work, we develop parallel implementations of key kernels to evaluate the merits of using GPUs as the baseband signal processor. We show that the digital processing subsystem for a 75 Mbps LTE base station can be built using two NVIDIA GTX 680 GPUs with power consumption of 188 W.

Original languageEnglish (US)
Title of host publicationIEEE Workshop on Signal Processing Systems, SiPS: Design and Implementation
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages219-224
Number of pages6
ISBN (Print)9781467362382
StatePublished - 2013
Event2013 IEEE Workshop on Signal Processing Systems, SiPS 2013 - Taipei, Taiwan, Province of China
Duration: Oct 16 2013Oct 18 2013

Other

Other2013 IEEE Workshop on Signal Processing Systems, SiPS 2013
CountryTaiwan, Province of China
CityTaipei
Period10/16/1310/18/13

Fingerprint

Graphics Processing Unit
Base stations
High level languages
Mobile Communication
Application specific integrated circuits
Parallel Implementation
Digital signal processing
Field Programmable Gate Array
Power Consumption
Workload
Field programmable gate arrays (FPGA)
Subsystem
Electric power utilization
Graphics processing unit
kernel
Resources
Evaluate
Communication
Costs

Keywords

  • Baseband processing
  • Graphics processing unit
  • LTE base station
  • Power efficient

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Signal Processing
  • Applied Mathematics
  • Hardware and Architecture

Cite this

Zheng, Q., Chen, Y., Dreslinski, R., Chakrabarti, C., Anastasopoulos, A., Mahlke, S., & Mudge, T. (2013). Architecting an LTE base station with graphics processing units. In IEEE Workshop on Signal Processing Systems, SiPS: Design and Implementation (pp. 219-224). [6674508] Institute of Electrical and Electronics Engineers Inc..

Architecting an LTE base station with graphics processing units. / Zheng, Q.; Chen, Y.; Dreslinski, R.; Chakrabarti, Chaitali; Anastasopoulos, A.; Mahlke, S.; Mudge, T.

IEEE Workshop on Signal Processing Systems, SiPS: Design and Implementation. Institute of Electrical and Electronics Engineers Inc., 2013. p. 219-224 6674508.

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

Zheng, Q, Chen, Y, Dreslinski, R, Chakrabarti, C, Anastasopoulos, A, Mahlke, S & Mudge, T 2013, Architecting an LTE base station with graphics processing units. in IEEE Workshop on Signal Processing Systems, SiPS: Design and Implementation., 6674508, Institute of Electrical and Electronics Engineers Inc., pp. 219-224, 2013 IEEE Workshop on Signal Processing Systems, SiPS 2013, Taipei, Taiwan, Province of China, 10/16/13.
Zheng Q, Chen Y, Dreslinski R, Chakrabarti C, Anastasopoulos A, Mahlke S et al. Architecting an LTE base station with graphics processing units. In IEEE Workshop on Signal Processing Systems, SiPS: Design and Implementation. Institute of Electrical and Electronics Engineers Inc. 2013. p. 219-224. 6674508
Zheng, Q. ; Chen, Y. ; Dreslinski, R. ; Chakrabarti, Chaitali ; Anastasopoulos, A. ; Mahlke, S. ; Mudge, T. / Architecting an LTE base station with graphics processing units. IEEE Workshop on Signal Processing Systems, SiPS: Design and Implementation. Institute of Electrical and Electronics Engineers Inc., 2013. pp. 219-224
@inproceedings{7920b924cd244151bea77adcc155f3f9,
title = "Architecting an LTE base station with graphics processing units",
abstract = "Due to the rapid growth of mobile communication, wireless base stations are becoming a significant consumer of computational resources. Historically, base stations have been built from ASICs, DSP processors, or FPGAs. This paper studies the feasibility of building wireless base stations from commercial graphics processing units (GPUs). GPUs are attractive because they are widely used massively parallel devices that can be programmed in a high level language. Base station workloads are highly parallel, making GPUs a potential candidate for a cost effective programmable solution. In this work, we develop parallel implementations of key kernels to evaluate the merits of using GPUs as the baseband signal processor. We show that the digital processing subsystem for a 75 Mbps LTE base station can be built using two NVIDIA GTX 680 GPUs with power consumption of 188 W.",
keywords = "Baseband processing, Graphics processing unit, LTE base station, Power efficient",
author = "Q. Zheng and Y. Chen and R. Dreslinski and Chaitali Chakrabarti and A. Anastasopoulos and S. Mahlke and T. Mudge",
year = "2013",
language = "English (US)",
isbn = "9781467362382",
pages = "219--224",
booktitle = "IEEE Workshop on Signal Processing Systems, SiPS: Design and Implementation",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - GEN

T1 - Architecting an LTE base station with graphics processing units

AU - Zheng, Q.

AU - Chen, Y.

AU - Dreslinski, R.

AU - Chakrabarti, Chaitali

AU - Anastasopoulos, A.

AU - Mahlke, S.

AU - Mudge, T.

PY - 2013

Y1 - 2013

N2 - Due to the rapid growth of mobile communication, wireless base stations are becoming a significant consumer of computational resources. Historically, base stations have been built from ASICs, DSP processors, or FPGAs. This paper studies the feasibility of building wireless base stations from commercial graphics processing units (GPUs). GPUs are attractive because they are widely used massively parallel devices that can be programmed in a high level language. Base station workloads are highly parallel, making GPUs a potential candidate for a cost effective programmable solution. In this work, we develop parallel implementations of key kernels to evaluate the merits of using GPUs as the baseband signal processor. We show that the digital processing subsystem for a 75 Mbps LTE base station can be built using two NVIDIA GTX 680 GPUs with power consumption of 188 W.

AB - Due to the rapid growth of mobile communication, wireless base stations are becoming a significant consumer of computational resources. Historically, base stations have been built from ASICs, DSP processors, or FPGAs. This paper studies the feasibility of building wireless base stations from commercial graphics processing units (GPUs). GPUs are attractive because they are widely used massively parallel devices that can be programmed in a high level language. Base station workloads are highly parallel, making GPUs a potential candidate for a cost effective programmable solution. In this work, we develop parallel implementations of key kernels to evaluate the merits of using GPUs as the baseband signal processor. We show that the digital processing subsystem for a 75 Mbps LTE base station can be built using two NVIDIA GTX 680 GPUs with power consumption of 188 W.

KW - Baseband processing

KW - Graphics processing unit

KW - LTE base station

KW - Power efficient

UR - http://www.scopus.com/inward/record.url?scp=84896463801&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84896463801&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:84896463801

SN - 9781467362382

SP - 219

EP - 224

BT - IEEE Workshop on Signal Processing Systems, SiPS: Design and Implementation

PB - Institute of Electrical and Electronics Engineers Inc.

ER -