A tool for modeling and simulation of approximate computing in hardware

• Grant number: 17/08015-8
• Support Opportunities: Scholarships in Brazil - Master
• Start date: October 01, 2017
• End date: February 28, 2019
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computer Systems
• Principal Investigator: Lucas Francisco Wanner
• Grantee: Isaías Bittencourt Felzmann
• Host Institution: Instituto de Computação (IC). Universidade Estadual de Campinas (UNICAMP). Campinas , SP, Brazil

Abstract

Approximate computing is a set of techniques that explore the generation of potentially incorrect or inaccurate results, when obtaining the most accurate result is not necessary or efficient for the given application. Approximate computing techniques have emerged in hardware, algorithms and applications development. Recent research shows that getting an approximate, but sufficient for the application, result is often less costly - specially in terms of energy consumption - than getting an accurate result. Despite the substantial interest in this research field, there is a lack of tools to support approximate hardware systems development. In this project, we propose a tool to explore approximate architectures. This tool, based in the ArchC architecture description language, will allow the exploitation of processors and systems where the result of hardware operations is approximated. These approximations will be derived from both the development of hardware components where the results are deterministic and approximated and the variation of parameters that are critical for the result reliability, such as temperature and source voltage. The tool will be launched as an opensource module, and will be verified using approximation-tolerant software as case studies, pointing design choices and the trade-off between quality of result and energy consumption. (AU)

Scientific publications

(References retrieved automatically from Web of Science and SciELO through information on FAPESP grants and their corresponding numbers as mentioned in the publications by the authors)

FELZMANN, ISAIAS B.; SUSIN, MATHEUS M.; DUENHA, LIANA; AZEVEDO, RODOLFO; WANNER, LUCAS F.; IEEE. ADeLe: Rapid Architectural Simulation for Approximate Hardware. 2018 30TH INTERNATIONAL SYMPOSIUM ON COMPUTER ARCHITECTURE AND HIGH PERFORMANCE COMPUTING (SBAC-PAD 2018), v. N/A, p. 8-pg., 2018-01-01. (17/08015-8)

FELZMANN, ISAIAS B.; FABRICIO FILHO, JOAO; AZEVEDO, RODOLFO; WANNER, LUCAS F.; BIANCHINI, CD; DESOUZA, PSL; DEBARROS, COF; FERREIRA, RAC. Impact of memory approximation on energy efficiency. 2018 SYMPOSIUM ON HIGH PERFORMANCE COMPUTING SYSTEMS (WSCAD 2018), v. N/A, p. 8-pg., 2018-01-01. (17/08015-8)

FELZMANN, ISAIAS B.; SUSIN, MATHEUS M.; DUENHA, LIANA; AZEVEDO, RODOLFO; WANNER, LUCAS F.. ADeLe: A description language for approximate hardware. FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE, v. 102, p. 245-258, JAN 2020. (17/08015-8)

Academic Publications

(References retrieved automatically from State of São Paulo Research Institutions)

FELZMANN, Isaías Bittencourt. Uma ferramenta para modelagem e simulação de computação aproximada em hardware. 2019. Master's Dissertation - Universidade Estadual de Campinas (UNICAMP). Instituto de Computação Campinas, SP.