Newsgroups: comp.lang.smalltalk
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!howland.reston.ans.net!cs.utexas.edu!news.ti.com!ticipa!clw
From: clw@peregrin.works.ti.com (Chris Winemiller)
Subject: Slopstones/Smopstones for Digitalk's Visual Smalltalk
Message-ID: <1994Dec28.065748.7695@ticipa.pac.sc.ti.com>
Sender: clw@ticipa.pac.sc.ti.com (Chris Winemiller)
Organization: None
Date: Wed, 28 Dec 1994 06:57:48 GMT
Lines: 106


I thought I'd post the Slopstone and Smopstone benchmark results for
Digitalk's new Visual Smalltalk 3.0.  For comparison I've also posted
the results from DT's Smalltalk/V Win32 2.0.2.

My machine:
486DX/33, 16M memory.

Summary:            Slopstones  Smopstones
Visual ST 3.0       0.63525403  1.36473884
V Win32 2.0.2       0.43958151  1.39511321

Looks like Visual ST improved 44.5% over V Win32 in Slopstones, but
decreased slightly (2.18%) in Smopstones.

Here are the full statistics ---

FOR VISUAL SMALLTALK 3.0:


Starting benchmarks with repetition count = 16000.

1000s    time    1000s of
itera-   sec-    iterations   slop-
tions    onds    per sec      stones   explanation

3808     0.71    5363      0.81263338     add integers
 544     2.69     202      0.84262701     add floats
 960     2.09     459      0.52078248     access strings
 320     0.71     451      1.28040973     create objects
 160     2.91      55      0.51385811     copy objects
 480     1.81     265      0.6239844      perform selectors
 896     4.34     206      0.28515416     evaluate blocks

 640  1.81460609  352      0.63525403     geometric mean

Benchmarks complete.


Starting benchmarks with repetition count = 1.

time in    smop-
seconds    stones    explanation

2.14     1.47523364     generating fractonaccis
1.1      1.02090909     generating primes
0.99     1.1020202      generating and parsing streams
1.76     1.75625        generating strings
1.6      0.729375       forming sets
2.69     1.91040892     sorting strings
2.58     2.17093023     sorcerer's apprentice

1.72552358     1.36473884     geometric mean

Benchmarks complete.

FOR V WIN32 2.0.2:

"Result from running SlopstoneBenchmark new runBenchmark."

Starting benchmarks with repetition count = 16000.

1000s    time    1000s of
itera-   sec-    iterations   slop-
tions    onds    per sec      stones   explanation

3808     1.1     3462     0.52451791     add integers
 544     3.9      139     0.58119658     add floats
 960     2.19     438     0.49700245     access strings
 320     1.38     232     0.65876153     create objects
 160     3.29      49     0.45450672     copy objects
 480     2.8      171     0.40336134     perform selectors
 896     7.14     125     0.173329       evaluate blocks

 640  2.62234831  244     0.43958151     geometric mean

Benchmarks complete.

"Result from running SmopstoneBenchmark new runBenchmark."


Starting benchmarks with repetition count = 1.

time in    smop-
seconds    stones    explanation

2.14     1.47523364     generating fractonaccis
1.16     0.96810345     generating primes
1.16     0.94051724     generating and parsing streams
2.25     1.37377778     generating strings
0.83     1.4060241      forming sets
2.64     1.94659091     sorting strings
2.75     2.03672727     sorcerer's apprentice

1.68795552     1.39511321     geometric mean

Benchmarks complete.


Regards,
Chris
==============================================================
Chris Winemiller               Internet: clw@works.ti.com
Disclaimer: I do not speak for TI.
==============================================================

