Theoretical Analysis and Evaluation of NoCs with Weighted Round-Robin Arbitration

Mandal, Sumit K.; Tong, Jie; Ayoub, Raid; Kishinevsky, Michael; Abousamra, Ahmed; Ogras, Ümit Y.

doi:10.1109/iccad51958.2021.9643448

Cited by 4 publications

(2 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…More recently, in order to represent packet flow in a NoC as an open feed-forward queuing network, Bhaskar and Venkatesh [12] proposed a mathematical model, with 10 and 15% of errors for NoCs with 36 routers and 64 routers respectively. Weighted Round-Robin arbitration has been proposed for priority based NoCs [13].…”

Section: Related Workmentioning

confidence: 99%

“…We can see that the error remains below 2% until reaching the saturation, where a maximum error of 12% is found. It is important to note that the increase in error when the network gets highly congested is common to all analytical models [7], [10]- [13]. This is the trade-off for extremely short computation times compared with very long simulation times (see Section IV-D).…”

Section: B Parameters Exploration With Synthetic Trafficsmentioning

confidence: 99%

See 1 more Smart Citation

Analytical Model for Performance Evaluation of Token-Passing-Based WiNoCs

Krayem,

Sosa,

Killian

et al. 2023

IEEE Des. Test

View full text Add to dashboard Cite

Recent advances in technology integration have introduced new on-chip interconnects, such as wireless Network-on-Chips (NoCs), making the design space too large to be efficiently explored with time-consuming standard simulators. In this paper, we propose an analytical model based on queuing theory to evaluate the latency of manycore architecture interconnects. We consider a hybrid interconnection that utilizes electrical and wireless NoCs for both intra-and inter-cluster communications. The results demonstrate that our proposed model significantly reduces the simulation execution time by up to 500× while maintaining an error rate of less than 5% compared to the Noxim cycle-accurate simulator.

show abstract

Section: Related Workmentioning

confidence: 99%