A VLSI architecture for concurrent data structures
A VLSI architecture for concurrent data structures
Deadlock-Free Message Routing in Multiprocessor Interconnection Networks
IEEE Transactions on Computers
An Adaptive and Fault Tolerant Wormhole Routing Strategy for k-ary n-cubes
IEEE Transactions on Computers
A New Theory of Deadlock-Free Adaptive Routing in Wormhole Networks
IEEE Transactions on Parallel and Distributed Systems
A Cost and Speed Model for k-ary n-Cube Wormhole Routers
IEEE Transactions on Parallel and Distributed Systems
Interconnection Networks: An Engineering Approach
Interconnection Networks: An Engineering Approach
k -ary n -trees: High Performance Networks for Massively Parallel Architectures
IPPS '97 Proceedings of the 11th International Symposium on Parallel Processing
Performance Evaluation of Adaptive Routing Algorithms for k-ary-n-cubes
PCRCW '94 Proceedings of the First International Workshop on Parallel Computer Routing and Communication
Networks on chip
An Energy-Efficient Network-on-Chip for a Heterogeneous Tiled Reconfigurable Systems-on-Chip
DSD '04 Proceedings of the Digital System Design, EUROMICRO Systems
Exploiting ECC Redundancy to Minimize Crosstalk Impact
IEEE Design & Test
| Hi-index | 0.00 |
In Networks-on-Chip (NoC) designs, crosstalk noise has become a serious issue which may cause the communication channel unreliable. The crosstalk problem can be mitigated by wide spacing of serial lines. However, the wider spacing of serial lines will reduce the number of the lines, thus reduce the data throughput. In this paper, a new fully adaptive multi-path routing (MPR) scheme is proposed to maximize the data throughput by utilizing multiple paths for concurrent data transmission. For the proposed MPR algorithm, two transport models are considered: the full-wire-bank transport model (FM) and the half-wire-bank transport model (HM). Theoretical analysis shows that the MPR scheme under both FM and HM achieves improvement in data throughput when single pair of nodes are in communication. When multiple pairs of nodes are in communication, simulation results demonstrate that the MPR scheme under FM significantly improves the normalized accepted traffic and throughput as well as average message latency than the YX routing algorithm in most network loads.