Last updated 3 September 2009.
We have graphed the performance of several implementations of the R6RS on a set of R6RS benchmarks collected by Will Clinger. These R6RS benchmarks include improved versions of most Gambit benchmarks (collected by Marc Feeley) and also include new benchmarks that test some of the more important features that were introduced by the R6RS.
Please keep in mind that the results of benchmarking should not be taken too seriously.
The R6RS-conforming Genuine R6RS benchmarks consist of portable R6RS top-level programs. We have timings for three very different machines:
As of this writing, there appear to be at least ten systems that implement enough of the R6RS to make benchmarking worthwhile. We benchmarked these seven:
We have not benchmarked Common Larceny, IronScheme, or Chez Scheme.
The numbers shown are for a single run on a machine with no other users, and give the elapsed time in seconds for a single run on a machine with no other users. The bar graphs show relative performance. Longer is better.
Where no timing is shown for a system, the system was unable to run the benchmark correctly. We believe that all missing timings reflect a bug or implementation restriction in the system that was unable to run the benchmark.