Performance Studies of Multiple-Packet Multistage Cube Networks and Comparison to Circuit Switching
نویسندگان
چکیده
This paper extends the modeling and analysis of packet switched multistage cube intercoimection networks to include the use of multiple-packet message formats. A multiplepacket message format is needed within an interconnection network when the message to be transmitted exceeds the network's packet size. The packet offset time is introduced and used to gauge the transmission rate differential between the network interchange boxes and the system PEs. Based on assumed network and system operating characteristics, optimum network performance offset values are given. A methodology for evaluating the relative performance of circuit and packet switching transmission modes is developed. For an assumed set of network and system parameters, results indicate that, in contrast to inter-computer networks, the performance of packet switching networks is better than that of circuit switching networks for long message lengths.
منابع مشابه
PERFORMANCE AND IMPLEMENTATION OF 4x4 SWITCHING NODES IN AN INTERCONNECTION NETWORK FOR PASM
Design issues for the multistage Generalized Cube network are discussed in this paper. An analysis of the merits of 2-input/2-output interchange boxes versus 4-input/4-output crossbars for interconnection network implementation is made. The cost and performance of each network for the two switching node alternatives are examined. Discussion of the sui tabi l i ty of each approach for VLSI imple...
متن کاملFault Location in Distributed Control Interconnection Networks
One class of networks suitable for use in parallel processing systems is the multistage cube network. Unfortunately, the cube network is not fault tolerant and any single failure within the network can prevent some sourcedestination communications. Cube networks with "extra" stages can be constructed that permit faults to be bypassed — providing the exact location of the fault is known. This pa...
متن کاملLow Latency on Chip Permutation Network for Multiprocessor System on Chip
This On-Chip Permutation Network for Multiprocessor System-On-Chip presents the silicon-proven design of a novel on-chip network to support guaranteed traffic permutation in multiprocessor system-on-chip applications the routing algorithm, the deflection routing is not energy-efficient due to the extra hops needed for deflected data transfer, compared to a minimal routing. Moreover, the deflect...
متن کاملRandomly-Wired Multistage Networks
Randomly-wired multistage networks have recently been shown to outperform traditional multistage networks in three respects. First, they have fast deterministic packet-switching and circuit-switching algorithms for routing permutations. Second, they are nonblocking, and there are on-line algorithms for establishing new connections in them, even if many requests for connections are made simultan...
متن کاملSwanet: a Novel Self-routed Wavelength-addressable Optical Switching Network 1
In this paper, we introduce a novel self-routed wavelength-addressable switching network (SWANET) that provides wavelength-transparent optical data paths between end-points, conng-ured by wavelength-coded optical signals. The network is based on a new scheme for encoding destination addresses using a sequence of wavelengths. This allows large networks to be constructed using a moderate number o...
متن کامل