earticle

논문검색

Novel Self-Compacting Buffer Schemes to Improve Performance of Systems with Network on Chip

초록

영어

In this paper, we present three novel input buffer schemes for systems with network on chip. These proposed buffer schemes are based on a self-compacting buffer, and can provide larger available buffer space per physical channel for communicating applications. These schemes outperform existing approaches. DAMQshr has similar performance using only sixty three percent of the buffer size that is used in traditional implementation for NoCs. DAMQmin provides an excellent technique to optimize buffer management, and provides a good throughput when the network has a larger load. In addition, they also have better utilization of the available buffer space.

목차

Abstract
 1. Introduction
 2. Background
 3. The Novel Buffer Schemes
  3.1. DAMQ with Reserved Space for all Virtual Channels Scheme
  3.2. DAMQ with Minimum Reserved Space for Virtual Channels Scheme
  3.3. Shared DAMQ Buffer Scheme
 4. Performance Evaluation
  4.1. Simulated Network Configurations
  4.2. Performance Metrics
  4.3. Simulation Results
 5. Conclusions and Future Work
 Acknowledgements
 References

저자정보

  • Jin Liu College of Information Engineering, Shanghai Maritime University
  • Hongmin Ren College of Information Engineering, Shanghai Maritime University
  • José G. Delgado-Frias School of EECS, Washington State University
  • Jeong-Uk Kim Department of Energy Grid, Sangmyung University
  • Jin Wang Computer and Software School, Nanjing University of Information Science & Technology

참고문헌

자료제공 : 네이버학술정보

    함께 이용한 논문

      ※ 원문제공기관과의 협약기간이 종료되어 열람이 제한될 수 있습니다.

      0개의 논문이 장바구니에 담겼습니다.