Revised from the original 1996 version for Netscape 2.0
These will run under Safari 4.0 and iPhone OS 3.0.
Warning: These are very small benchmarks that have poor correlation
with real benchmarks such as SPECs.
- primes will calculate and display prime numbers from 5 to 1010.
PRIMES.BAS took about 25 seconds to run on an Apple ][+.
This benchmark is divide intensive.
- pgap will calculate primes up to 15729 and display the largest
gap (prime void) between primes found so far. The pgap benchmark
compiled in C took about 3 seconds of run time on a VAX 11/780.
- sieve uses a sieve algorithm to count prime numbers up to 2*8192.
Byte Magazine ran a couple articles about this sieve benchmarks in
'82 and '83. This benchmark is array access intensive.
- fib(20) recursively calculates the 20th fibonachi number.
The fib benchmark took about a half a second to run in compiled C
on a VAX 11/780. This benchmark is function call intensive.
- tak is another small recursive (Takeuchi function)
benchmark that was popular for evaluating Lisp system performance
in the early '80s. The tak benchmark took a around 4 seconds to run
under InterLisp on a VAX 11/780.
- mb100 calculates the number of points in a 100x100 window of
the Mandelbrot set with a depth of > 90.
This benchmark is floating point intensive.
Mail to Ron Nicholson at
Copyright © 1995,1996,2009 Ronald H. Nicholson, Jr. ALL RIGHTS RESERVED.
Last Modified June 21, 2009