Stack data management for Limited Local Memory (LLM) multi-core processors

Ke Bai, Aviral Shrivastava, Saleel Kudchadker

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

18 Citations (Scopus)

Abstract

Limited Local Memory (LLM) architectures are power-efficient, scalable memory multi-core architectures, in which cores have a scratch-pad like local memory that is software controlled. Any data transfers between the main memory and the local memory must be explicitly present as Direct Memory Access (DMA) commands in the application. Stack data management of the cores is an important problem in LLM architecture, and our previous work outlined a promising scheme for that [1]. In this paper, we improve the previous approach, and now can i) manage limitless stack data, ii) increase the applicability of stack management, and iii) perform stack management with smaller footprint on the local memory. We demonstrate these by executing benchmarks from the MiBench suite on the IBM Cell processor.

Original languageEnglish (US)
Title of host publicationProceedings of the International Conference on Application-Specific Systems, Architectures and Processors
Pages231-234
Number of pages4
DOIs
StatePublished - 2011
Event22nd IEEE International Conference on Application-Specific Systems, Architectures and Processors, ASAP 2011 - Santa Monica, CA, United States
Duration: Sep 11 2011Sep 14 2011

Other

Other22nd IEEE International Conference on Application-Specific Systems, Architectures and Processors, ASAP 2011
CountryUnited States
CitySanta Monica, CA
Period9/11/119/14/11

Fingerprint

Information management
Data storage equipment
Memory architecture
Data transfer

Keywords

  • embedded systems
  • IBM Cell
  • local memory
  • MPI
  • multi-core processor
  • scratch pad memory
  • Stack

ASJC Scopus subject areas

  • Hardware and Architecture
  • Computer Networks and Communications

Cite this

Bai, K., Shrivastava, A., & Kudchadker, S. (2011). Stack data management for Limited Local Memory (LLM) multi-core processors. In Proceedings of the International Conference on Application-Specific Systems, Architectures and Processors (pp. 231-234). [6043275] https://doi.org/10.1109/ASAP.2011.6043275

Stack data management for Limited Local Memory (LLM) multi-core processors. / Bai, Ke; Shrivastava, Aviral; Kudchadker, Saleel.

Proceedings of the International Conference on Application-Specific Systems, Architectures and Processors. 2011. p. 231-234 6043275.

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

Bai, K, Shrivastava, A & Kudchadker, S 2011, Stack data management for Limited Local Memory (LLM) multi-core processors. in Proceedings of the International Conference on Application-Specific Systems, Architectures and Processors., 6043275, pp. 231-234, 22nd IEEE International Conference on Application-Specific Systems, Architectures and Processors, ASAP 2011, Santa Monica, CA, United States, 9/11/11. https://doi.org/10.1109/ASAP.2011.6043275
Bai K, Shrivastava A, Kudchadker S. Stack data management for Limited Local Memory (LLM) multi-core processors. In Proceedings of the International Conference on Application-Specific Systems, Architectures and Processors. 2011. p. 231-234. 6043275 https://doi.org/10.1109/ASAP.2011.6043275
Bai, Ke ; Shrivastava, Aviral ; Kudchadker, Saleel. / Stack data management for Limited Local Memory (LLM) multi-core processors. Proceedings of the International Conference on Application-Specific Systems, Architectures and Processors. 2011. pp. 231-234
@inproceedings{41956662bc494b41be9a3496f71cef7c,
title = "Stack data management for Limited Local Memory (LLM) multi-core processors",
abstract = "Limited Local Memory (LLM) architectures are power-efficient, scalable memory multi-core architectures, in which cores have a scratch-pad like local memory that is software controlled. Any data transfers between the main memory and the local memory must be explicitly present as Direct Memory Access (DMA) commands in the application. Stack data management of the cores is an important problem in LLM architecture, and our previous work outlined a promising scheme for that [1]. In this paper, we improve the previous approach, and now can i) manage limitless stack data, ii) increase the applicability of stack management, and iii) perform stack management with smaller footprint on the local memory. We demonstrate these by executing benchmarks from the MiBench suite on the IBM Cell processor.",
keywords = "embedded systems, IBM Cell, local memory, MPI, multi-core processor, scratch pad memory, Stack",
author = "Ke Bai and Aviral Shrivastava and Saleel Kudchadker",
year = "2011",
doi = "10.1109/ASAP.2011.6043275",
language = "English (US)",
isbn = "9781457712920",
pages = "231--234",
booktitle = "Proceedings of the International Conference on Application-Specific Systems, Architectures and Processors",

}

TY - GEN

T1 - Stack data management for Limited Local Memory (LLM) multi-core processors

AU - Bai, Ke

AU - Shrivastava, Aviral

AU - Kudchadker, Saleel

PY - 2011

Y1 - 2011

N2 - Limited Local Memory (LLM) architectures are power-efficient, scalable memory multi-core architectures, in which cores have a scratch-pad like local memory that is software controlled. Any data transfers between the main memory and the local memory must be explicitly present as Direct Memory Access (DMA) commands in the application. Stack data management of the cores is an important problem in LLM architecture, and our previous work outlined a promising scheme for that [1]. In this paper, we improve the previous approach, and now can i) manage limitless stack data, ii) increase the applicability of stack management, and iii) perform stack management with smaller footprint on the local memory. We demonstrate these by executing benchmarks from the MiBench suite on the IBM Cell processor.

AB - Limited Local Memory (LLM) architectures are power-efficient, scalable memory multi-core architectures, in which cores have a scratch-pad like local memory that is software controlled. Any data transfers between the main memory and the local memory must be explicitly present as Direct Memory Access (DMA) commands in the application. Stack data management of the cores is an important problem in LLM architecture, and our previous work outlined a promising scheme for that [1]. In this paper, we improve the previous approach, and now can i) manage limitless stack data, ii) increase the applicability of stack management, and iii) perform stack management with smaller footprint on the local memory. We demonstrate these by executing benchmarks from the MiBench suite on the IBM Cell processor.

KW - embedded systems

KW - IBM Cell

KW - local memory

KW - MPI

KW - multi-core processor

KW - scratch pad memory

KW - Stack

UR - http://www.scopus.com/inward/record.url?scp=80055085209&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=80055085209&partnerID=8YFLogxK

U2 - 10.1109/ASAP.2011.6043275

DO - 10.1109/ASAP.2011.6043275

M3 - Conference contribution

SN - 9781457712920

SP - 231

EP - 234

BT - Proceedings of the International Conference on Application-Specific Systems, Architectures and Processors

ER -