Use este identificador para citar ou linkar para este item:
Título: AxRAM: a lightweight implicit interface for approximate data access
Autor(es): Fabrício Filho, João
Felzmann, Isaías Bittencourt
Azevedo, Rodolfo Jardim de
Wanner, Lucas Francisco
Palavras-chave: Tolerância a falha (Computadores)
Arquitetura de computador
Energia - Consumo
Tolerância a falha (Computadores)
Computer architecture
Computer science
Energy consumption
Data do documento: Dez-2020
Câmpus: Campo Mourao
Citação: FABRÍCIO FILHO, João; FELZMANN, Isaías B.; AZEVEDO, Rodolfo J.; WANNER, Lucas F. AxRAM: a lightweight implicit interface for approximate data access. Future Generation Computer Systems, v. 113, n. 1, p. 556-570, 2020. DOI: Disponível em: Acesso em: 28 dez. 2020.
Abstract: Approximate memories expose data elements to errors in order to improve energy efficiency. For a large fraction of data, these errors are inconsequential or lead only to small losses in application output quality. Nevertheless, for some critical data, errors may lead to execution flow crashes, resulting in non-produced outputs and wasted computational and energy resources. Thus, these techniques require some level of control over approximations to generate acceptable outputs and, consequently, to maximize the energy benefits. Many proposed interfaces for approximate memories rely on burdensome instrumentation of the program or on user annotations to protect critical data. We present AxRAM, a lightweight interface for approximate data that avoids crashes without user annotations. AxRAM relies on a memory with configurable reliability levels and protects from errors critical data regions commonly found on a number of applications. Furthermore, our interface implements a resilient addressing scheme that reduces invalid data accesses that lead to execution crashes. In an embedded computing scenario with a dual-VDD SRAM, our implementation of AxRAM reduces 51% of the execution crashes compared to an unprotected approximate memory, resulting in energy savings for 9 out of 12 profiled applications.
Descrição: O arquivo disponibilizado é a versão Preprint do artigo publicado na Elsevier.
ISSN: 1872-7115
Aparece nas coleções:PCS - Artigos

Arquivos associados a este item:
Arquivo Descrição TamanhoFormato 
  Disponível a partir de 2022-07-17
1,22 MBAdobe PDFThumbnail

Este item está licenciada sob uma Licença Creative Commons Creative Commons