Open Computing History Project

A visiting research assistant spent the summer running period benchmark programs across six home computers from the project's collection. The aim was to see how well the published magazine benchmarks of the time hold up when repeated on real hardware that is now more than 25 years old.

All six machines ran the same short set of BASIC programs: loops, string handling, floating point arithmetic and screen output. We timed each run with a stopwatch, as the original reviewers did, and repeated each one five times. In most cases our results came within a few percent of the published figures. The largest difference came from one machine whose results depended heavily on whether the display was enabled during the run.

The full results and the program listings are now in the archive. We plan to repeat the exercise next year with the machines that are currently waiting for repair.