fond
Model Checking Contest 2020
10th edition, Paris, France, June 23, 2020
smart compared to other tools («Surprise» models, QuasiLiveness)
Last Updated
Jun 28, 2020

Introduction

This page presents how smart do cope efficiently with the QuasiLiveness examination face to the other participating tools. In this page, we consider «Surprise» models.

The next sections will show chart comparing performances in terms of both memory and execution time.The x-axis corresponds to the challenging tool where the y-axes represents smart' performances. Thus, points below the diagonal of a chart denote comparisons favorables to the tool while others corresponds to situations where the challenging tool performs better.

You might also find plots out of the range that denote the case were at least one tool could not answer appropriately (error, time-out, could not compute or did not competed).

smart versus GreatSPN

Some statistics are displayed below, based on 422 runs (211 for smart and 211 for GreatSPN, so there are 211 plots on each of the two charts). Each execution was allowed 1 hour and 16 GByte of memory. Then performance charts comparing smart to GreatSPN are shown (you may click on one graph to enlarge it).

Statistics on the executions
  smart GreatSPN Both tools   smart GreatSPN
All computed OK 0 26 102   Smallest Memory Footprint
smart = GreatSPN 0 Times tool wins 0 128
smart > GreatSPN 0   Shortest Execution Time
smart < GreatSPN 0 Times tool wins 12 116
Do not compete 20 0 0
Error detected 0 0 0  
Cannot Compute + Time-out 20 14 69


On the chart below, denote cases where the two tools did computed all results without error, denote cases where the two tool did computed the same number of values (but not al values in the examination), denote cases where smart computed more values than GreatSPN, denote cases where smart computed less values than GreatSPN, denote the cases where at least one tool did not competed, denote the cases where at least one tool computed a bad value and denote the cases where at least one tool stated it could not compute a result or timed-out.

smart wins when points are below the diagonal, GreatSPN wins when points are above the diagonal.

memory chart time chart

smart versus Tapaal

Some statistics are displayed below, based on 422 runs (211 for smart and 211 for Tapaal, so there are 211 plots on each of the two charts). Each execution was allowed 1 hour and 16 GByte of memory. Then performance charts comparing smart to Tapaal are shown (you may click on one graph to enlarge it).

Statistics on the executions
  smart Tapaal Both tools   smart Tapaal
All computed OK 13 43 89   Smallest Memory Footprint
smart = Tapaal 0 Times tool wins 13 132
smart > Tapaal 0   Shortest Execution Time
smart < Tapaal 0 Times tool wins 39 106
Do not compete 20 0 0
Error detected 0 0 0  
Cannot Compute + Time-out 43 33 46


On the chart below, denote cases where the two tools did computed all results without error, denote cases where the two tool did computed the same number of values (but not al values in the examination), denote cases where smart computed more values than Tapaal, denote cases where smart computed less values than Tapaal, denote the cases where at least one tool did not competed, denote the cases where at least one tool computed a bad value and denote the cases where at least one tool stated it could not compute a result or timed-out.

smart wins when points are below the diagonal, Tapaal wins when points are above the diagonal.

memory chart time chart