From: Myriam M Abramson
Subject: ACL 5.0
Date:
Message-ID: <MABRAMSO.98Oct29111641@osf1.gmu.edu>
Personally, I'm very impressed with the new version. But
it seems that the performance went down in some aspects.
Any reason why? Does anybody know what the different
functions in the Gabriel benchmarks measure?
Those timings were taken on an SGI R4000, Irix 6.2.
Thanks,
(declaim (optimize (speed 3)))
acl 4.3 acl 4.3 opt acl 5.0 acl 5.0 opt
------- ----------- ------- -----------
BOYER 0.370 BOYER 0.370 BOYER 1.450 BOYER 1.410
BROWSE 0.590 BROWSE 0.480 BROWSE 1.390 BROWSE 1.250
CTAK 0.120 CTAK 0.140 CTAK 0.100 CTAK 0.100
DDERIV 0.130 DDERIV 0.130 DDERIV 0.530 DDERIV 0.500
DERIV 0.110 DERIV 0.120 DERIV 0.470 DERIV 0.470
DESTRU-MOD 0.093 DESTRU-MOD 0.060 DESTRU-MOD 0.135 DESTRU-MOD 0.100
DESTRU 0.093 DESTRU 0.062 DESTRU 0.135 DESTRU 0.100
DIV2 0.130 DIV2 0.130 DIV2 0.480 DIV2 0.480
FFT-MOD 3.040 FFT-MOD 0.050 FFT-MOD 4.360 FFT-MOD 0.050
FFT 2.920 FFT 0.050 FFT 4.350 FFT 0.050
FPRINT 0.282 FPRINT 0.338 FPRINT 0.312 FPRINT 0.312
FREAD 0.157 FREAD 0.150 FREAD 0.165 FREAD 0.170
FRPOLY 1.700 FRPOLY 1.550 FRPOLY 2.570 FRPOLY 2.580
PUZZLE-MOD 2.360 PUZZLE-MOD 0.370 PUZZLE-MOD 2.240 PUZZLE-MOD 0.300
PUZZLE 2.080 PUZZLE 0.260 PUZZLE 2.190 PUZZLE 0.210
STAK 0.162 STAK 0.130 STAK 0.233 STAK 0.210
TAK-MOD 0.220 TAK-MOD 0.130 TAK-MOD 0.170 TAK-MOD 0.120
TAK 0.210 TAK 0.130 TAK 0.170 TAK 0.120
TAKL 0.120 TAKL 0.120 TAKL 0.100 TAKL 0.100
TAKR 0.083 TAKR 0.060 TAKR 0.072 TAKR 0.038
TPRINT 0.475 TPRINT 0.480 TPRINT 0.548 TPRINT 0.543
TRAVERSE 3.460 TRAVERSE 1.520 TRAVERSE 3.370 TRAVERSE 1.270
TRIANG-MOD 18.650 TRIANG-MOD 16.82 TRIANG-MOD 22.030 TRIANG-MOD 20.89
TRIANG 22.270 TRIANG 3.380 TRIANG 20.000 TRIANG 2.900
--
myriam