620.omnetpp_s
SPEC CPU2017 Benchmark Description

Benchmark Name

620.omnetpp_s

Benchmark Authors

Jerome Rousselot
CSEM SA

Andras Varga
andras [at] omnetpp.org

Rudolf Horning
rudi [at] omnetpp.org

Benchmark Program General Category

Discrete Event Simulation

Benchmark Description

The benchmark performs discrete event simulation of a large 10 gigabit Ethernet network. The simulation is based on the OMNeT++ discrete event simulation system ([1]www.omnetpp.org), a generic and open simulation framework. OMNeT++'s primary application area is the simulation of communication networks, but its generic and flexible architecture allows for its use in other areas such as the simulation of IT systems, queueing networks, hardware architectures or business processes as well.

The ten gigabit Ethernet network model used in this benchmark is publicly available as part of the INET Framework from the address given in the References.

For the reference workload, the simulated 10 Gbps network models a large network comprised of six backbone switches, eight small LANs, twenty medium LANs and twelve large LANs.

Input Description

The structure of hosts is described in NED files (NED is the network description language of OMNeT++.) Operation of the physical layer models, MAC, traffic generator, etc. are in C++.

Several model parameters are specified in the omnetpp.ini file.

Parameters modified for the different workloads are as follows:

Output Description

The model computes summary statistics during the simulation execution to minimize disk space usage and facilitate the results analysis. The node that receives the traffic generated by all network nodes computes for each source node several latency parameters (among which: mean, median, minimum and maximum). In addition, the same latency parameters are also computed for all received packets. At the routing layer, the number of packets sent, received and forwarded, as well as route statistics are also generated. This is typically used to track down route establishment difficulties. At the MAC layer, each node keeps track of the number of frames sent and received, along with backoff statistics. This type of information is useful to identify network congestion.

Programming Language

C++

Known portability issues

None

Sources and Licensing

The benchmark is licensed directly to SPEC by OpenSim Ltd. Note: therefore, source code references to other terms under which the program may be available are not relevant for the SPEC CPU version. It uses a version of Mersenne Twister under BSD license; for details, please see SPEC CPU2017 Licenses.

References

Last updated: $Date: 2017-05-01 13:34:29 -0400 (Mon, 01 May 2017) $