Tightness and computation assessment of worst-case delay bounds in wormhole networks-on-chip
Author(s) -
Frédéric Giroudot,
Ahlem Mifdaoui
Publication year - 2019
Publication title -
open archive toulouse archive ouverte (university of toulouse)
Language(s) - English
Resource type - Conference proceedings
ISBN - 978-1-4503-7223-7
DOI - 10.1145/3356401.3356408
Subject(s) - wormhole , computer science , network calculus , computation , network on a chip , channel (broadcasting) , virtual channel , computer network , parallel computing , distributed computing , algorithm , quality of service , physics , quantum mechanics
This paper addresses the problem of worst-case timing analysis in wormhole Networks-On-Chip (NoCs). We consider our previous work [5] for computing maximum delay bounds using Network Calculus, called the Buffer-Aware Worst-case Timing Analysis (BATA). The latter allows the computation of delay bounds for a large panel of wormhole NoCs, e.g., handling priority-sharing, Virtual Channel Sharing and buffer backpressure.In this paper, we provide further insights into the tightness and computation issues of the worst-case delay bounds yielded by BATA. Our assessment shows that the gap between the computed delay bounds and the worst-case simulation results is reasonably small (70% tighness on average). Furthermore, BATA provides good delay bounds for medium-scale configurations within less than one hour. Finally, we evaluate the yielded improvements with BATA for a realistic use-case against a recent state-of-the-art approach. This evaluation shows the applicability of BATA under more general assumptions and the impact of such a feature on the tightness and computation time.
Accelerating Research
Robert Robinson Avenue,
Oxford Science Park, Oxford
OX4 4GP, United Kingdom
Address
John Eccles HouseRobert Robinson Avenue,
Oxford Science Park, Oxford
OX4 4GP, United Kingdom