Toward Ideal On-Chip Communication Using Express Virtual Channels

Kumar, Amit; Peh, Li-Shiuan; Kundu, Partha; Jha, Niraj K.

doi:10.1109/mm.2008.18

Cited by 79 publications

(31 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Because of the complexity of the dynamic EVC approach [113], in this work, we adopt a 3-hop static EVC network; this provides a better trade off between performance and design complexity. We also define a region as a 4×4 mesh whose boundary is made up of the EVC source/sink nodes.…”

Section: Motivation For Adding Hub Routersmentioning

confidence: 99%

“…Then, the express path arbiter will grant the requests from hub latches to use the output link with a higher priority than the normal paths. In order to avoid the starvation of flits in normal buffers and registers, the arbiter will grant the requests from normal paths after serving the EVC and hub latches for n consecutive cycles as in [113]. In this work, n is set to be 30 according to our simulation results.…”

Section: Router Implementationmentioning

confidence: 99%

“…Indeed, enhancing the NoC architecture with additional flexibility is appealing since it can offer more opportunities to minimize the impact of traffic congestion by bypassing the intermediate router pipeline stages for some packets at run time [113]. At the same time, a routing strategy, which fully utilizes the characteristics of both the underlying communication infrastructure and the NoC traffic, is equally important since the routing decision made at the run time is critical in optimizing the NoC performance and resources utilization [37,114].…”

Section: Introductionmentioning

confidence: 99%

“…Our proposed NoC architecture exploits the following two capabilities, namely: i) self reconfigurable bidirectional channels (BiNoC) [100,102] and ii) static express virtual channels (EVCs) [113] with regional hub routers, to dynamically create short-cut paths for transmitting packets. In order to exploit the added flexibility provided by the proposed architecture, we also propose a new fitness-based adaptive routing algorithm for optimizing the network performance.…”

Section: Introductionmentioning

confidence: 99%

“…In [113] Previous adaptive routing algorithms are mainly designed for unidirectional NoCs. In [114], a neighbors-on-path (NoP) selection strategy is proposed to make each node routing selection based on the condition of the nodes adjacent to the neighbors.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

High performance network-on-chips (NoCs) design : performance modeling, routing algorithm and architecture optimization

Qian¹

View full text Add to dashboard Cite

TABLE OF CONTENTS Title Page i Authorization Page ii Signature Page iiiAcknowledgments iv Table of Contents v List of Figures ix List of Tables xiiiAbstract xiv Therefore, fast and accurate analytical models for NoC-based multicore performance evaluation are strongly desired to better explore the design space. For this purpose, we propose a machine learning based latency regression model to evaluate the NoC designs with respect to different configurations before the system is built or taped-out. Then, for high performance NoC designs, we tackle one of the most important problems, i.e., the routing algorithms design with different design constraints and objectives. For avoiding temperature hotspots, a thermal-aware routing algorithm is proposed to achieve an even temperature profile for application-specific Network-on-chips (NoCs). For improving the reliability, a routing algorithm to achieve maximum performance under fault is proposed. Finally, in the architecture level, we propose two new NoC structures using bi-directional links for the performance optimization. In particular, we propose a flit-level speedup scheme to enhance the network-on-chip(NoC) performance utilizing bidirectional channels. In addition to the traditional efforts on allowing flits of different packets using the idling internal and external bandwidth of the bidirectional channel, our proposed flit-level speedup scheme also allows flits within the same packet to be transmitted simultaneously on the bi-directional channel.We also propose a flexible NoC architecture which takes advantage of a dynamic distributed routing algorithm and improves the NoC communication performance with minimal energy overhead. This proposed NoC architecture exploits the selfreconfigurable bidirectional channels to increase the effective bandwidth and uses express virtual paths, as well as localized hub routers, to bypass some intermediate nodes at run time in the network. From the simulation results on both synthetic traffic and real workload traces, significantly performance improvement in terms of latency and throughput can be achieved.

show abstract

Section: Motivation For Adding Hub Routersmentioning

confidence: 99%

Section: Router Implementationmentioning

confidence: 99%