Expand this Topic clickable element to expand a topic
Skip to content
Optica Publishing Group

High-Performance Optical Packet Switching Network for Green Data Centers

Open Access Open Access

Abstract

A five-year-long R&D program (2011~) funded by NICT, Japan will be outlined, which aims at developing optical packet switching (OPS) network for green data center (DC) accommodating huge number of servers. The consortium includes four institutes; NTT Laboratories, NEC Corp., Osaka University, and Kyushu University. In the intra-DC network, hybrid optoelectronic packet routers (HOERs) are interconnected with 100Gbps (25Gbps x 4 wavelengths) links. A novel “Express path” is proposed to enables the flow control. The traffic, which is part of a flow, can be bypassed the intermediate nodes and directly forwarded to the destination through the express path. On the other hand, an individual packet is processed and forwarded by the HOERs. OpenFlow will be introduced to enable the flow control in a centralized fashion. At last but not the least, a drastic reduction of the power consumption is expected due to a substantial offload from the packet forwarding by using the proposed the express path as well as low-power photonic hardware of HOER.

© 2012 Optical Society of America

PDF Article
More Like This
1.3 μm SDN-enabled Optical Packet Switch Architecture for High Performance and Programmable Data Center Network

W. Miao, F. Agraz, H. de Waardt, S. Spadaro, H.J.S. Dorren, and N. Calabretta
Th2A.66 Optical Fiber Communication Conference (OFC) 2015

Scalable and Low Latency Optical Packet Switching Architectures for High Performance Data Center Networks

Nicola Calabretta, Wang Miao, Stefano Di Lucente, Jun Luo, and Harm Dorren
PW4B.1 Photonics in Switching (PS) 2014

On the Latencies in a Hybrid Optical Packet Switching Network in Data Center

Artur Minakhmetov, Archana Nagarajan, Luigi Iannone, and Cédric Ware
W2A.21 Optical Fiber Communication Conference (OFC) 2019

Select as filters


Select Topics Cancel
© Copyright 2024 | Optica Publishing Group. All rights reserved, including rights for text and data mining and training of artificial technologies or similar technologies.