Inter-cluster communication in VLIW architectures

A.S. Terechko, H. Corporaal

Research output: Contribution to journalArticleAcademicpeer-review

12 Citations (Scopus)

Abstract

The traditional VLIW (very long instruction word) architecture with a single register file does not scale up well to address growing performance demands on embedded media processors. However, splitting a VLIW processor in smaller clusters, which are comprised of function units fully connected to local register files, can significantly improve VLSI implementation characteristics of the processor, such as speed, energy consumption, and area. In our paper we reveal that achieving the best characteristics of a clustered VLIW requires a thorough selection of an Inter-cluster Communication (ICC) model, which is the way clustering is exposed in the Instruction Set Architecture. For our study we, first, define a taxonomy of ICC models including copy operations, dedicated issue slots, extended operands, extended results, and multicast. Evaluation of the execution time of the models requires both the dynamic cycle count and clock period. We developed an advanced instruction scheduler for all the five ICC models in order to quantify the dynamic cycle counts of our multimedia C benchmarks. To assess the clock period of the ICC models we designed and laid out VLIW datapaths using the RTL hardware descriptions derived from a deeply pipelined commercial TriMedia processor. In contrast to prior art, our research shows that fully distributed register file architectures (with eight clusters in our study) often underperform compared to moderately clustered machines with two or four clusters because of explosion of the cycle count overhead in the former. Among the evaluated ICC models, performance of the copy operation model, popular both in academia and industry, is severely limited by the copy operations hampering scheduling of regular operations in high ILP (instruction-level parallelism) code. The dedicated issue slots model combats this limitation by dedicating extra VLIW issue slots purely for ICC, reaching the highest 1.74 execution time speedup relative to the unicluster. Furthermore, our VLSI experiments show that the lowest area and energy consumption of 42 and 57% relative to the unicluster, respectively, are achieved by the extended operands model, which, nevertheless, provides higher performance than the copy operation model.
Original languageEnglish
Pages (from-to)11-1/38
JournalACM Transactions on Architecture and Code Optimization
Volume4
Issue number2
DOIs
Publication statusPublished - 2007

Fingerprint

Very long instruction word architecture
Communication
Clocks
Energy utilization
Taxonomies
Explosions

Cite this

@article{bbc725d3f8614a8694dadefa3e8b00f8,
title = "Inter-cluster communication in VLIW architectures",
abstract = "The traditional VLIW (very long instruction word) architecture with a single register file does not scale up well to address growing performance demands on embedded media processors. However, splitting a VLIW processor in smaller clusters, which are comprised of function units fully connected to local register files, can significantly improve VLSI implementation characteristics of the processor, such as speed, energy consumption, and area. In our paper we reveal that achieving the best characteristics of a clustered VLIW requires a thorough selection of an Inter-cluster Communication (ICC) model, which is the way clustering is exposed in the Instruction Set Architecture. For our study we, first, define a taxonomy of ICC models including copy operations, dedicated issue slots, extended operands, extended results, and multicast. Evaluation of the execution time of the models requires both the dynamic cycle count and clock period. We developed an advanced instruction scheduler for all the five ICC models in order to quantify the dynamic cycle counts of our multimedia C benchmarks. To assess the clock period of the ICC models we designed and laid out VLIW datapaths using the RTL hardware descriptions derived from a deeply pipelined commercial TriMedia processor. In contrast to prior art, our research shows that fully distributed register file architectures (with eight clusters in our study) often underperform compared to moderately clustered machines with two or four clusters because of explosion of the cycle count overhead in the former. Among the evaluated ICC models, performance of the copy operation model, popular both in academia and industry, is severely limited by the copy operations hampering scheduling of regular operations in high ILP (instruction-level parallelism) code. The dedicated issue slots model combats this limitation by dedicating extra VLIW issue slots purely for ICC, reaching the highest 1.74 execution time speedup relative to the unicluster. Furthermore, our VLSI experiments show that the lowest area and energy consumption of 42 and 57% relative to the unicluster, respectively, are achieved by the extended operands model, which, nevertheless, provides higher performance than the copy operation model.",
author = "A.S. Terechko and H. Corporaal",
year = "2007",
doi = "10.1145/1250727.1250731",
language = "English",
volume = "4",
pages = "11--1/38",
journal = "ACM Transactions on Architecture and Code Optimization",
issn = "1544-3566",
publisher = "Association for Computing Machinery, Inc",
number = "2",

}

Inter-cluster communication in VLIW architectures. / Terechko, A.S.; Corporaal, H.

In: ACM Transactions on Architecture and Code Optimization, Vol. 4, No. 2, 2007, p. 11-1/38.

Research output: Contribution to journalArticleAcademicpeer-review

TY - JOUR

T1 - Inter-cluster communication in VLIW architectures

AU - Terechko, A.S.

AU - Corporaal, H.

PY - 2007

Y1 - 2007

N2 - The traditional VLIW (very long instruction word) architecture with a single register file does not scale up well to address growing performance demands on embedded media processors. However, splitting a VLIW processor in smaller clusters, which are comprised of function units fully connected to local register files, can significantly improve VLSI implementation characteristics of the processor, such as speed, energy consumption, and area. In our paper we reveal that achieving the best characteristics of a clustered VLIW requires a thorough selection of an Inter-cluster Communication (ICC) model, which is the way clustering is exposed in the Instruction Set Architecture. For our study we, first, define a taxonomy of ICC models including copy operations, dedicated issue slots, extended operands, extended results, and multicast. Evaluation of the execution time of the models requires both the dynamic cycle count and clock period. We developed an advanced instruction scheduler for all the five ICC models in order to quantify the dynamic cycle counts of our multimedia C benchmarks. To assess the clock period of the ICC models we designed and laid out VLIW datapaths using the RTL hardware descriptions derived from a deeply pipelined commercial TriMedia processor. In contrast to prior art, our research shows that fully distributed register file architectures (with eight clusters in our study) often underperform compared to moderately clustered machines with two or four clusters because of explosion of the cycle count overhead in the former. Among the evaluated ICC models, performance of the copy operation model, popular both in academia and industry, is severely limited by the copy operations hampering scheduling of regular operations in high ILP (instruction-level parallelism) code. The dedicated issue slots model combats this limitation by dedicating extra VLIW issue slots purely for ICC, reaching the highest 1.74 execution time speedup relative to the unicluster. Furthermore, our VLSI experiments show that the lowest area and energy consumption of 42 and 57% relative to the unicluster, respectively, are achieved by the extended operands model, which, nevertheless, provides higher performance than the copy operation model.

AB - The traditional VLIW (very long instruction word) architecture with a single register file does not scale up well to address growing performance demands on embedded media processors. However, splitting a VLIW processor in smaller clusters, which are comprised of function units fully connected to local register files, can significantly improve VLSI implementation characteristics of the processor, such as speed, energy consumption, and area. In our paper we reveal that achieving the best characteristics of a clustered VLIW requires a thorough selection of an Inter-cluster Communication (ICC) model, which is the way clustering is exposed in the Instruction Set Architecture. For our study we, first, define a taxonomy of ICC models including copy operations, dedicated issue slots, extended operands, extended results, and multicast. Evaluation of the execution time of the models requires both the dynamic cycle count and clock period. We developed an advanced instruction scheduler for all the five ICC models in order to quantify the dynamic cycle counts of our multimedia C benchmarks. To assess the clock period of the ICC models we designed and laid out VLIW datapaths using the RTL hardware descriptions derived from a deeply pipelined commercial TriMedia processor. In contrast to prior art, our research shows that fully distributed register file architectures (with eight clusters in our study) often underperform compared to moderately clustered machines with two or four clusters because of explosion of the cycle count overhead in the former. Among the evaluated ICC models, performance of the copy operation model, popular both in academia and industry, is severely limited by the copy operations hampering scheduling of regular operations in high ILP (instruction-level parallelism) code. The dedicated issue slots model combats this limitation by dedicating extra VLIW issue slots purely for ICC, reaching the highest 1.74 execution time speedup relative to the unicluster. Furthermore, our VLSI experiments show that the lowest area and energy consumption of 42 and 57% relative to the unicluster, respectively, are achieved by the extended operands model, which, nevertheless, provides higher performance than the copy operation model.

U2 - 10.1145/1250727.1250731

DO - 10.1145/1250727.1250731

M3 - Article

VL - 4

SP - 11-1/38

JO - ACM Transactions on Architecture and Code Optimization

JF - ACM Transactions on Architecture and Code Optimization

SN - 1544-3566

IS - 2

ER -