» Implementation and Evaluation of a Dynamically Routed Processor Operand Network

| Sorted by Date | Classified by Publication Type | Classified by Research Category |

Paul Gratz, Karthikeyan Sankaralingam, Heather Hanson, Premkishore Shivakumar, Robert McDonald, Stephen Keckler, and Doug Burger. Implementation and Evaluation of a Dynamically Routed Processor Operand Network. In Proceedings of the 1st ACM/IEEE International Symposium on Networks-on-Chip, May 2007.

Download

[HTML]

Abstract

Microarchitecturally integrated on-chip networks, or micronets, are candidates to replace busses for processor component interconnect in future processor designs. For micronets, tight coupling between processor microarchitecture and network architecture is one of the keys to improving processor performance. This paper presents the design, implementation and evaluation of the TRIPS operand network (OPN). The TRIPS OPN is a 5times5, dynamically routed, 2D mesh micronet that is integrated into the TRIPS microprocessor core. The TRIPS OPN is used for operand passing, register file I/O, and primary memory system I/O. We discuss in detail the OPN design, including the unique features that arise from its integration with the processor core, such as its connection to the execution unit's wakeup pipeline and its in flight mis-speculated traffic removal. We then evaluate the performance of the network under synthetic and realistic loads. Finally, we assess the processor performance implications of OPN design decisions with respect to the end-to-end latency of OPN packets and the OPN's bandwidth.

Additional Information

This is a test of the extra info broadcasting system.

BibTeX

 @inproceedings{nocs2007:tripsocn,
  author =       {Paul Gratz and Karthikeyan Sankaralingam and Heather Hanson and Premkishore Shivakumar and Robert McDonald and Stephen Keckler and Doug Burger},
  title =        "{Implementation and Evaluation of a Dynamically Routed Processor Operand Network}",
  booktitle =    "{Proceedings of the 1st ACM/IEEE International Symposium on Networks-on-Chip}",
  year =         2007,
  month =        {May},
  abstract = {
 Microarchitecturally integrated on-chip networks, or micronets, are candidates to replace busses for processor component interconnect in future processor designs. For micronets, tight coupling between processor microarchitecture and network architecture is one of the keys to improving processor performance. This paper presents the design, implementation and evaluation of the TRIPS operand network (OPN). The TRIPS OPN is a 5times5, dynamically routed, 2D mesh micronet that is integrated into the TRIPS microprocessor core. The TRIPS OPN is used for operand passing, register file I/O, and primary memory system I/O. We discuss in detail the OPN design, including the unique features that arise from its integration with the processor core, such as its connection to the execution unit's wakeup pipeline and its in flight mis-speculated traffic removal. We then evaluate the performance of the network under synthetic and realistic loads. Finally, we assess the processor performance implications of OPN design decisions with respect to the end-to-end latency of OPN packets and the OPN's bandwidth.
 },
  bib_dl = "http://ieeexplore.ieee.org/xpl/freeabs_all.jsp?isnumber=4208982&arnumber=4208990&count=46&index=7",
  bib_pubtype = {Refereed Conference},
  bib_rescat = {Architecture,Hardware},
  bib_extra_info = {This is a test of the extra info broadcasting system.}
 }

Generated by bib.pl (written by Patrick Riley ) on Sat Jul 15, 2017 14:55:44 time=1207019082


Page last modified on October 23, 2017