TY - GEN
T1 - A new dynamic bandwidth re-allocation technique in optically interconnected high-performance computing systems
AU - Kodi, Avinash Karanth
AU - Louri, Ahmed
PY - 2006
Y1 - 2006
N2 - As bit rates increase, optical interconnects based high-performance computing (HPC) systems improve performance by increasing the available bandwidth (using wavelength-division multiplexing (WDM) and space-division multiplexing (SDM)) and decreasing power dissipation as compared to traditional electrical interconnects. While static allocation of wavelengths (channels) in optical interconnects provide every node with equal opportunity for communication, it can lead to network congestion for non-uniform traffic patterns. In this paper, we propose an opto-electronic interconnect for designing a flexible, high-bandwidth, low-latency, dynamically reconfigurable architecture for scalable HPC systems. Re configurability is realized by monitoring traffic intensities, and implementing dynamic bandwidth re-allocation (DBR) technique that adapts to changes in communication patterns. We propose a DBR technique - Lock-Step (LS) that balances the load on each communication channel based on past utilization. Simulation results indicate that the reconfigured architecture shows 40% increased throughput and and 20% reduced network latency as compared to HPC electrical networks.
AB - As bit rates increase, optical interconnects based high-performance computing (HPC) systems improve performance by increasing the available bandwidth (using wavelength-division multiplexing (WDM) and space-division multiplexing (SDM)) and decreasing power dissipation as compared to traditional electrical interconnects. While static allocation of wavelengths (channels) in optical interconnects provide every node with equal opportunity for communication, it can lead to network congestion for non-uniform traffic patterns. In this paper, we propose an opto-electronic interconnect for designing a flexible, high-bandwidth, low-latency, dynamically reconfigurable architecture for scalable HPC systems. Re configurability is realized by monitoring traffic intensities, and implementing dynamic bandwidth re-allocation (DBR) technique that adapts to changes in communication patterns. We propose a DBR technique - Lock-Step (LS) that balances the load on each communication channel based on past utilization. Simulation results indicate that the reconfigured architecture shows 40% increased throughput and and 20% reduced network latency as compared to HPC electrical networks.
UR - http://www.scopus.com/inward/record.url?scp=34547423246&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=34547423246&partnerID=8YFLogxK
U2 - 10.1109/HOTI.2006.6
DO - 10.1109/HOTI.2006.6
M3 - Conference contribution
AN - SCOPUS:34547423246
SN - 0769526543
SN - 9780769526546
T3 - Proceedings - Symposium on the High Performance Interconnects, Hot Interconnects
SP - 31
EP - 36
BT - Proceedings - 14TH IEEE Symposium on High Performance Interconnects, Hot Interconnects, HotI-14
T2 - 14TH IEEE Symposium on High Performance Interconnects, Hot Interconnects, HotI-14
Y2 - 23 August 2006 through 25 August 2006
ER -