[gambit-list] Draft note about some scheme benchmarks

Bradley Lucier lucier at math.purdue.edu
Wed Nov 29 16:10:46 EST 2017


On 11/29/2017 10:31 AM, Sven Hartrumpf wrote:

> Many Scheme implementations perform differently for different ABIs:
> x86-64 is often _slower_ than x32 

Thanks for your note.  That is an interesting point.

Right now I'm trying to understand what various benchmarks are trying to 
measure and whether they're succeeding.

Partly, I'm concluding that certain of the r7rs-benchmarks have small 
changes that partially obscure what I think they're trying to measure.

Brad



More information about the Gambit-list mailing list