|
Analysis time comparison: SFL versus Pos x SFL
[Page last updated on 2001/02/10 18:04:22.]
The analysis time comparison (goal dependent analysis, without structural information, differences report)
Benchmark | Base domain | Enhanced domain |
CoVer.pl | 3.12 | 4.45 |
LeanTaP.pl | 6.34 | 9.01 |
action.pl | 14.92 | 25.19 |
aircraft.pl | 0.38 | 0.50 |
ann.pl | 0.65 | 0.76 |
boyer_wamcc.pl | 0.23 | 0.33 |
bryant.pl | 0.53 | 0.73 |
chat80.pl | 444.90 | 474.91 |
chat_parser.pl | 165.72 | 197.60 |
circan.pl | 34.49 | 34.25 |
dposan.pl | 0.52 | 0.68 |
ezan.pl | 29.67 | 34.03 |
gnup-1.1.0_pl2wam.pl | 37.10 | 40.00 |
lojban.pl | 27.00 | 66.48 |
lptp-1.05.pl | 79.68 | 129.74 |
lptp-1.06.pl | 79.63 | 129.93 |
mgpl4.pl | 0.88 | 1.03 |
mgtp-g-I-temp.pl | 0.32 | 0.46 |
mm.pl | 3.09 | 3.22 |
music.pl | 48.45 | 46.89 |
poker.pl | 1.12 | 1.31 |
press.pl | 0.43 | 0.55 |
quotan.pl | 2.03 | 2.58 |
sdda.pl | 2.70 | 3.07 |
semigroup.pl | 2.40 | 2.55 |
simple_analyzer.pl | 5.52 | 8.66 |
sprftp.pl.unsound | 132.16 | 132.00 |
strips.pl | 13.97 | 17.95 |
synth.pl | 71.60 | 74.51 |
trs.pl | 343.61 | 348.55 |
unify.pl | 0.63 | 0.74 |
vhdl97_parser.pl | 28.89 | 30.43 |
Column headers legenda
Benchmark |
Base domain |
Enhanced domain |
The name of the benchmark. |
The analysis fixpoint computation time
for the base domain, in seconds. |
The analysis fixpoint computation time
for the enhanced domain, in seconds. |
Background colors legenda
For those looking at this page through color-enabled devices:
Background color |
Meaning |
Performance improvement |
Table cells with a light green background indicate
that the investigated technique led to
a performance improvement. |
Performance loss |
Table cells with a red background indicate
that the investigated technique led to
a performance loss. |
Same time |
Table cells with a normal background indicate
that the two analyses terminated in the same time,
up to the fixed time threshold of 0.1 seconds. |
Unknown |
Table cells with a light brown background indicate
that the time comparison was not possible because both
analyses timed-out. |
|