Multi-dimensional optimization for approximate near-threshold computing

Jing WANG; Wei-wei LIANG; Yue-hua NIU; Lan GAO; Wei-gong ZHANG

doi:10.1631/FITEE.2000089

Your Location：

Home >

Browse articles >

Multi-dimensional optimization for approximate near-threshold computing

Regular Papers | Updated：2022-05-19

- Multi-dimensional optimization for approximate near-threshold computing
- 支持近似计算的近阈值系统多目标优化
- Frontiers of Information Technology & Electronic Engineering Vol. 21, Issue 10, Pages: 1426-1441(2020)
- Affiliations：
  
  College of Information Engineering, Capital Normal University, Beijing 100056, China
  Institution of Spacecraft System Engineering China Academy of Space Technology, Beijing 100094, China
  Beijing Advanced Innovation Center for Imaging Theory and Technology, Beijing 100048, China
- Author bio：
  
  [ "Jing WANG, E-mail: jwang@cnu.edu.cn" ]
  Wei-gong ZHANG, E-mail:zwg771@cnu.edu.cn
- Funds：
- DOI：10.1631/FITEE.2000089
  CLC： TP302.1
- Received：26 February 2020，
  
  Revised：2020-;9-29，
  
  Published：2020-10
- Accepted：
Scan QR Code
Jing WANG, Wei-wei LIANG, Yue-hua NIU, et al. Multi-dimensional optimization for approximate near-threshold computing[J]. Frontiers of Information Technology & Electronic Engineering, 2020, 21(10): 1426-1441.
DOI：

Jing WANG, Wei-wei LIANG, Yue-hua NIU, et al. Multi-dimensional optimization for approximate near-threshold computing[J]. Frontiers of Information Technology & Electronic Engineering, 2020, 21(10): 1426-1441. DOI： 10.1631/FITEE.2000089.

摘要

登纳德缩放定律的失效使计算机系统面临功耗和利用率双重挑战。让晶体管在近阈值电压附近工作，能够有效解决能耗墙问题。然而，电压降低会引发错误，导致可靠性问题。若在解决电压降低带来的副作用的同时确保系统完全正确，又会额外减损系统性能，增加能耗。由此可见，计算机系统设计的目标已从简单的性能优化发展到多目标综合优化。本文提出一种通过有效识别系统最佳配置实现性能、能耗和可靠性的综合优化方法。设计了输出精度预测器、性能预测器和功耗预测器，分别预测不同系统配置下的精度、性能和功耗。其中输出质量预测器采用软硬件协同的故障注入平台，分析近阈值电压导致的错误对输出精度的影响。采用多目标优化动态规划模型，基于所设计的输出精度预测器、性能预测器和功耗预测器，选择系统最佳的电压和近似级别。实验结果显示本文方案在能效性下降10%的情况下将输出精度提高28%，从而实现平均20%的精度、功耗和性能的综合优化。

Abstract

The demise of Dennardos scaling has created both power and utilization wall challenges for computer systems. As transistors operating in the near-threshold region are able to obtain flexible trade-offs between power and performance

it is regarded as an alternative solution to the scaling challenge. A reduction in supply voltage will nevertheless generate significant reliability challenges

while maintaining an error-free system that generates high costs in both performance and energy consumption. The main purpose of research on computer architecture has therefore shifted from performance improvement to complex multi-objective optimization. In this paper

we propose a three-dimensional optimization approach which can effectively identify the best system configuration to establish a balance among performance

energy

and reliability. We use a dynamic programming algorithm to determine the proper voltage and approximate level based on three predictors: system performance

energy consumption

and output quality. We propose an output quality predictor which uses a hardware/software co-design fault injection platform to evaluate the impact of the error on output quality under near-threshold computing (NTC). Evaluation results demonstrate that our approach can lead to a 28% improvement in output quality with a 10% drop in overall energy efficiency; this translates to an approximately 20% average improvement in accuracy

power

and performance.

关键词

Keywords

references

O Azizi , , , A Mahesri , , , BC Lee , , , 等 . . Energy-performance tradeoffs in processor architecture and circuit design: a marginal cost analysis . . ACM SIGARCH Comput Arch News , , 2010 . . 38 ( ( 3 ): ): 26 - - 36 . . DOI: 10.1145/1816038.1815967 http://doi.org/10.1145/1816038.1815967 . .

TE Carlson , , , W Heirman , , , L Eeckhout . . Sniper: exploring the level of abstraction for scalable and accurate parallel multi-core simulation . . Proc Int Conf for High Performance Computing, Networking, Storage and Analysis , , 2011 . . p.1 - - 12 . . DOI: 10.1145/2063384.2063454 http://doi.org/10.1145/2063384.2063454 . .

VK Chippa , , , ST Chakradhar , , , K Roy , , , 等 . . Analysis and characterization of inherent application resilience for approximate computing . . 50 th ACM/EDAC/IEEE Design Automation Conf , , 2013 . . p.1 - - 9 . . DOI: 10.1145/2463209.2488873 http://doi.org/10.1145/2463209.2488873 . .

S Das , , , D Blaauw , , , D Bull , , , 等 . . Addressing design margins through error-tolerant circuits . . 46 th ACM/IEEE Design Automation Conf , , 2009 . . p.11 - - 12 . . DOI: 10.1145/1629911.1629917 http://doi.org/10.1145/1629911.1629917 . .

H Esmaeilzadeh , , , A Sampson , , , L Ceze , , , 等 . . Neural acceleration for general-purpose approximate programs . . 45 th Annual IEEE/ACM Int Symp on Microarchitecture , , 2012 . . p.449 - - 460 . . DOI: 10.1109/MICRO.2012.48 http://doi.org/10.1109/MICRO.2012.48 . .

K Ferreira , , , J Stearley , , , JH Laros , , , 等 . . Evaluating the viability of process replication reliability for exascale systems . . Proc Int Conf for High Performance Computing, Networking, Storage and Analysis , , 2011 . . p.1 - - 12 . . DOI: 10.1145/2063384.2063443 http://doi.org/10.1145/2063384.2063443 . .

B Grigorian , , , N Farahpour , , , G Reinman . . BRAINIAC: bringing reliable accuracy into neurally-implemented approximate computing . . IEEE 21 st Int Symp on High Performance Computer Architecture , , 2015 . . p.615 - - 626 . . DOI: 10.1109/HPCA.2015.7056067 http://doi.org/10.1109/HPCA.2015.7056067 . .

V Gupta , , , D Mohapatra , , , SP Park , , , 等 . . IMPACT: IMPrecise adders for low-power approximate computing . . IEEE/ACM Int Symp on Low Power Electronics and Design , , 2011 . . p.409 - - 414 . . DOI: 10.1109/ISLPED.2011.5993675 http://doi.org/10.1109/ISLPED.2011.5993675 . .

KH Huang , , , JA Abraham . . Algorithm-based fault tolerance for matrix operations . . IEEE Trans Comput , , 1984 . . C-33 ( ( 6 ): ): 518 - - 528 . . DOI: 10.1109/TC.1984.1676475 http://doi.org/10.1109/TC.1984.1676475 . .

UR Karpuzcu , , , KB Kolluru , , , NS Kim , , , 等 . . VARIUS- NTV: a microarchitectural model to capture the increased sensitivity of manycores to process variations at near- threshold voltages . . IEEE/IFIP Int Conf on Dependable Systems and Networks , , 2012 . . p.1 - - 11 . . DOI: 10.1109/DSN.2012.6263951 http://doi.org/10.1109/DSN.2012.6263951 . .

H Kaul , , , M Anders , , , S Hsu , , , 等 . . Near-threshold voltage (NTV) design-opportunities and challenges . . Proc 49 th Annual Design Automation Conf , , 2012 . . p.1149 - - 1154 . . DOI: 10.1145/2228360.2228572 http://doi.org/10.1145/2228360.2228572 . .

V Kozhikkottu , , , S Venkataramani , , , S Dey , , , 等 . . Variation tolerant design of a vector processor for recognition, mining and synthesis . . Proc Int Symp on Low Power Electronics and Design , , 2014 . . p.239 - - 244 . . DOI: 10.1145/2627369.2627636 http://doi.org/10.1145/2627369.2627636 . .

S Liu , , , K Pattabiraman , , , T Moscibroda , , , 等 . . Flikker: saving DRAM refresh-power through critical data partitioning . . Proc 16 th Int Conf on Architectural Support for Programming Languages and Operating Systems , , 2011 . . p.213 - - 224 . . DOI: 10.1145/1950365.1950391 http://doi.org/10.1145/1950365.1950391 . .

B Reagen , , , U Gupta , , , L Pentecost , , , 等 . . Ares: a framework for quantifying the resilience of deep neural networks . . Proc 55 th ACM/ESDA/IEEE Design Automation Conf , , 2018 . . p.1 - - 6 . . DOI: 10.1109/DAC.2018.8465834 http://doi.org/10.1109/DAC.2018.8465834 . .

M Samadi , , , DA Jamshidi , , , J Lee , , , 等 . . Paraprox: pattern- based approximation for data parallel applications . . Int Conf on Architectural Support for Programming Languages and Operating Systems , , 2014 . . p.35 - - 50 . . DOI: 10.1145/2541940.2541948 http://doi.org/10.1145/2541940.2541948 . .

A Sampson , , , A Baixo , , , B Ransford , , , 等 . . ACCEPT: a Programmer-Guided Compiler Framework for Practical Approximate Computing . . Technical Report No. UW-CSE-15-01, University of Washington, USA , , 2015 . . .

MH Santriaji , , , H Hoffmann . . GRAPE: minimizing energy for GPU applications with performance requirements . . 49 th Annual IEEE/ACM Int Symp on Microarchitecture , , 2016 . . p.1 - - 13 . . DOI: 10.1109/MICRO.2016.7783719 http://doi.org/10.1109/MICRO.2016.7783719 . .

A Shye , , , T Moseley , , , VJ Reddi , , , 等 . . Using process-level redundancy to exploit multiple cores for transient fault tolerance . . 37 th Annual IEEE/IFIP Int Conf on Dependable Systems and Networks , , 2007 . . p.297 - - 306 . . DOI: 10.1109/DSN.2007.98 http://doi.org/10.1109/DSN.2007.98 . .

S Sidiroglou-Douskos , , , S Misailovic , , , H Hoffmann , , , 等 . . Managing performance vs . . accuracy trade-offs with loop perforation. Proc 19 th ACM SIGSOFT Symp and 13 th European Conf on Foundations of Software Engineering , , 2011 . . p.124 - - 134 . . DOI: 10.1145/2025113.2025133 http://doi.org/10.1145/2025113.2025133 . .

C Silvano , , , G Palermo , , , S Xydis , , , 等 . . Voltage island management in near threshold manycore architectures to mitigate dark silicon . . Design, Automation & Test in Europe Conf & Exhibition , , 2014 . . p.1 - - 6 . . DOI: 10.7873/DATE.2014.214 http://doi.org/10.7873/DATE.2014.214 . .

W Song , , , S Mukhopadhyay , , , S Yalamanchili . . Architectural reliability: lifetime reliability characterization and management of many-core processors . . IEEE Comput Arch Lett , , 2015a . . 14 ( ( 2 ): ): 103 - - 106 . . DOI: 10.1109/LCA.2014.2340873 http://doi.org/10.1109/LCA.2014.2340873 . .

W Song , , , S Mukhopadhyay , , , S Yalamanchili . . Managing performance-reliability tradeoffs in multi-core processors . . IEEE Int Reliability Physics Symp , , 2015b . . p.p.3C.1.1 - - 3C.1.7 . . DOI: 10.1109/IRPS.2015.7112707 http://doi.org/10.1109/IRPS.2015.7112707 . .

M Sutherland , , , J San Miguel , , , N Enright Jerger . . http://www.eecg.toronto.edu/~enright/TexCacheApprox.pdf http://www.eecg.toronto.edu/~enright/TexCacheApprox.pdf , , Texture cache approximation on GPUs . . University of Toronto, Toronto, Canada , , 2015 . . .

R Tavakkoli-Moghaddam , , , J Safari , , , F Sassani . . Reliability optimization of series-parallel systems with a choice of redundancy strategies using a genetic algorithm . . Reliab Eng Syst Saf , , 2008 . . 93 ( ( 4 ): ): 550 - - 556 . . DOI: 10.1016/j.ress.2007.02.009 http://doi.org/10.1016/j.ress.2007.02.009 . .

R Teodorescu , , , J Torrellas . . Variation-aware application scheduling and power management for chip multiprocessors . . Int Symp on Computer Architecture , , 2008 . . p.363 - - 374 . . DOI: 10.1109/ISCA.2008.40 http://doi.org/10.1109/ISCA.2008.40 . .

Y Tian , , , Q Zhang , , , T Wang , , , 等 . . ApproxMA: approximate memory access for dynamic precision scaling . . Proc 25th Edition on Great Lakes Symp on VLSI , , 2015 . . p.337 - - 342 . . DOI: 10.1145/2742060.2743759 http://doi.org/10.1145/2742060.2743759 . .

R Venkatagiri , , , A Mahmoud , , , SKS Hari , , , 等 . . Approxilyzer: towards a systematic framework for instruction- level approximate computing and its application to hardware resiliency . . 49th Annual IEEE/ACM Int Symp on Microarchitecture , , 2016 . . p.1 - - 14 . . DOI: 10.1109/MICRO.2016.7783745 http://doi.org/10.1109/MICRO.2016.7783745 . .

L Wang , , , JA Rivers , , , MS Gupta , , , 等 . . Resilience and real-time constrained energy optimization in embedded processor systems . . 10th Workshop on Silicon Errors in Logic-System Effects , , 2014 . . .

L Wang , , , AJ Vega , , , A Buyuktosunoglu , , , 等 . . Power- efficient embedded processing with resilience and real- time constraints . . IEEE/ACM Int Symp on Low Power Electronics and Design , , 2015 . . p.231 - - 236 . . DOI: 10.1109/ISLPED.2015.7273519 http://doi.org/10.1109/ISLPED.2015.7273519 . .

HJ Wunderlich , , , C Braun , , , A Schll . . Pushing the limits: how fault tolerance extends the scope of approximate computing . . IEEE 22nd Int Symp on On-line Testing and Robust System Design , , 2016 . . p.133 - - 136 . . DOI: 10.1109/IOLTS.2016.7604686 http://doi.org/10.1109/IOLTS.2016.7604686 . .

A Yazdanbakhsh , , , D Mahajan , , , H Esmaeilzadeh , , , 等 . . AxBench: a multiplatform benchmark suite for approximate computing . . IEEE Des Test , , 2017 . . 34 ( ( 2 ): ): 60 - - 68 . . DOI: 10.1109/MDAT.2016.2630270 http://doi.org/10.1109/MDAT.2016.2630270 . .

Y Zhang , , , K Chakrabarty . . A unified approach for fault tolerance and dynamic power management in fixed- priority real-time embedded systems . . IEEE Trans Comput-Aid Des Int Circ Syst , , 2006 . . 25 ( ( 1 ): ): 111 - - 125 . . DOI: 10.1109/TCAD.2005.852657 http://doi.org/10.1109/TCAD.2005.852657 . .

BX Zhao , , , H Aydin , , , DK Zhu . . Reliability-aware dynamic voltage scaling for energy-constrained real-time embedded systems . . IEEE Int Conf on Computer Design , , 2008 . . p.633 - - 639 . . DOI: 10.1109/ICCD.2008.4751927 http://doi.org/10.1109/ICCD.2008.4751927 . .

LL Zhong . . BROAD: Bold and Reliable Online Approximate Computing Framework for Diverse Applications . . MS Thesis, University of Illinois at Urbana- Champaign, Urbana, Illinois, USA , , 2015 . . .

Views

Downloads

CSCD

Alert me when the article has been cited

Submit

Tools

Publicity Resources

A spaceborne advanced storage system for remote sensing microsatellites

Efficient and optimized approximate GDI full adders based on dynamic threshold CNTFETs for specific least significant bits

Related Author

Shilei TU

Huiquan WANG

Yue HUANG

Zhonghe JIN

Ayoub SADEGHI

Razieh GHASEMI

Hossein GHASEMIAN

Nabiollah SHIRI

Related Institution

Micro-Satellite Research Center, Zhejiang University

Key Laboratory of Micro-Nano Satellite Research

Department of Electrical Engineering, Shiraz Branch, Islamic Azad University

School of Electrical Engineering, Iran University of Science and Technology

Department of Electrical and Electronic Engineering, Shiraz University of Technology

Chat

Address：Zhejiang University Press, 148 Tianmushan Road, Hangzhou, China Postal code：310028
Tel：+86-571-88273162 Email：fitee@zju.edu.cn
It is recommended to read the content of this site in Chrome&IE9+. Please switch to extreme mode in browser 360.
Cookies We use cookies to help provide and enhance our service and tailor content. By continuing, you agree to the use of cookies.

⁰