On 3/24/2019 1:20 PM, Phil Nguyen wrote:
With Racket 7.2, the following program takes >13 seconds to run on
Windows, and <7 seconds on Linux either on Virtualbox on the same
machine, or native Linux on another machine with slightly
lower-powered processor:
|
#lang racket/base(define (fact n)(if(zero?n)1(*n (fact (-n 1)))))(time
(void(fact 100000)))
;Windowsnative,i7-7660U; cpu time:13610real time:13633gc
time:4459;Linuxon Virtualbox,i7-7660U; cpu time:6691real time:6706gc
time:1298;Linuxnative,i7-8500Y:; cpu time:6894real time:6882gc time:1129
|
While the difference is unlikely to matter in practice, given `fact 100000` is
a very large number, I'm curious what accounts for this difference? Is it some
big-integer library that Racket relies on?
I haven't paid attention to it recently, but in early 6.x versions I
observed very large differences in bytecode compilation speeds ... on
the order of 5 to 10 times with Linux (even on VM) the clear winner.
See the discussion here:
https://groups.google.com/forum/#!searchin/racket-users/neuner$20linux%7Csort:date/racket-users/iisjAKujpFQ/GJxa-PhHXRoJ
Execution times for a linked executable or an application consisting of
pre-compiled .zo files appeared to be roughly equivalent. Similarly,
starting a linked executable appeared roughly equivalent, but Windows
was 2..3 times slower than Linux at starting "racket <app>" [most
likely due to checking the .zo files].
Note however that I was not benchmarking but rather working on a rather
complicated web service, so I was judging execution speeds mainly from
timestamping in my application logs.
George
--
You received this message because you are subscribed to the Google Groups "Racket
Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
For more options, visit https://groups.google.com/d/optout.
