FOGSim overview

FOGSim is a modular simulator for interconnection networks, developed by the Department of Computer Science and Electronics in the University of Cantabria, Spain.

FOGSim models dragonfly interconnection networks using synthetic traffic patterns or traces of parallel applications as inputs. FOGSim results provides insight on network performance latency, throughput, traffic unfairness, response time to transient traffic, etc. FOGSim has been used for research on topology, routing, deadlock avoidance, and congestion avoidance mechanisms, among others.

Simulator description

Please see the wiki for a description of the simulator design and its basic use.

Developers and maintainers

FOGSim has been developed mainly by Marina García, Miguel Odriozola, Mariano Benito and Pablo Fuentes, under the supervision of Enrique Vallejo and Ramón Beivide. Additional contributions have been received from Cristóbal Camarero.

Current maintainers are Mariano Benito and Pablo Fuentes.

Publications using FOGSim

The following list of publications employed FOGSim to obtain performance results:

• On-the-Fly Adaptive Routing in High-Radix Hierarchical Networks. M. García, E. Vallejo, R. Beivide, M. Odriozola, C. Camarero, M. Valero, G. Rodríguez, J. Labarta and C. Minkenberg. International Conference on Parallel Processing (ICPP-41), 2012. (Best paper).
• Global Misrouting Policies in Two-level Hierarchical Networks. M. García, E. Vallejo, R. Beivide, M. Odriozola, C. Camarero, M. Valero, G. Rodriguez and J. Labarta. International Workshop on Interconnection Network Architecture: On-Chip, Multi-Chip (INA-OCMC), 2013.
• OFAR-CM: Efficient Dragonfly Networks with Simple Congestion Management. M. García, E. Vallejo, R. Beivide, M. Valero and G. Rodríguez. Symposium on High-Performance Interconnects (HOTI-21), 2013.
• Efficient Routing Mechanisms for Dragonfly Networks. M. García, E. Vallejo, R. Beivide, M. Odriozola and M. Valero. International Conference on Parallel Processing (ICPP-42), 2013.
• On the use of contention information for adaptive routing. P. Fuentes, E. Vallejo, M. García and R. Beivide. International Summer School on Advanced Computer Architecture and Compilation for Embedded Systems (ACACES-10), 2014.
• On-the-Fly Adaptive Routing for dragonfly interconnection networks. M. García, E. Vallejo, R. Beivide, C. Camarero, M. Valero, G. Rodríguez and C. Minkenberg. The Journal of Supercomputing, 2014.
• Contention-based Nonminimal Adaptive Routing in High-radix Networks. P. Fuentes, E. Vallejo, M. García, R. Beivide, G. Rodríguez, C. Minkenberg and M. Valero. International Parallel & Distributed Processing Symposium (IPDPS), 2015.
• Throughput Unfairness in Dragonfly Networks under Realistic Traffic Patterns. P. Fuentes, E. Vallejo, C. Camarero, R. Beivide, and M. Valero. 1st IEEE International Workshop on High-Performance Interconnection Networks Towards the Exascale and Big-Data Era (HiPINEB), 2015.
• On the Use of Commodity Ethernet Technology in Exascale HPC Systems. M. Benito, E. Vallejo, and R. Beivide. IEEE International Conference on High Performance Computing (HiPC), 2015.