[MLton-devel] benchmarks vs 20020410
Stephen Weeks
MLton@mlton.org
Sun, 25 Aug 2002 16:31:22 -0700
Here are the results of benchmarking my current working version
against 20020410. As before, it shows no cost of card marking. It
also shows some minor improvements and worsenings, but nothing
significant (the worst is life, which I just verified is a cache
effect).
MLton0 -- /usr/local/bin/mlton
MLton1 -- mlton -mark-cards false
MLton2 -- mlton -mark-cards true
compile time
benchmark MLton0 MLton1 MLton2
barnes-hut 1.5 1.5 1.5
boyer 3.1 3.3 3.4
checksum 0.4 0.4 0.4
count-graphs 1.0 1.0 1.0
DLXSimulator 2.8 3.0 3.0
fft 0.8 0.8 0.8
fib 0.3 0.3 0.3
hamlet 29.2 32.3 32.4
imp-for 0.4 0.4 0.4
knuth-bendix 1.2 1.3 1.3
lexgen 3.4 3.6 3.7
life 0.8 0.8 0.8
logic 1.7 1.8 1.8
mandelbrot 0.4 0.4 0.4
matrix-multiply 0.4 0.4 0.4
md5 0.7 0.7 0.7
merge 0.4 0.4 0.4
mlyacc 13.1 15.6 15.6
mpuz 0.5 0.5 0.5
nucleic 1.8 3.1 2.9
peek 0.7 0.7 0.7
psdes-random 0.4 0.4 0.4
ratio-regions 1.6 1.7 1.7
ray 2.5 2.3 2.3
raytrace 6.5 7.4 7.4
simple 4.7 4.8 4.9
smith-normal-form 5.5 5.4 4.8
tailfib 0.3 0.4 0.4
tak 0.4 0.4 0.3
tensor 1.9 2.0 2.0
tsp 0.8 0.9 0.9
tyan 2.2 2.4 2.4
vector-concat 0.4 0.4 0.4
vector-rev 0.4 0.4 0.4
vliw 8.1 8.5 8.5
wc-input1 1.1 1.0 1.1
wc-scanStream 1.1 1.0 1.0
zebra 3.8 4.2 4.2
zern 0.7 0.8 0.7
run time
benchmark MLton0 MLton1 MLton2
barnes-hut 1.7 1.8 1.9
boyer 2.4 2.6 2.7
checksum 1.6 1.6 1.6
count-graphs 1.7 1.8 1.9
DLXSimulator 11.1 10.2 10.8
fft 8.8 8.2 8.2
fib 1.7 1.6 1.5
hamlet 4.0 4.1 4.3
imp-for 3.5 3.7 3.7
knuth-bendix 3.4 3.4 3.4
lexgen 7.6 6.3 6.4
life 2.9 4.0 4.1
logic 8.7 9.0 9.7
mandelbrot 3.5 3.5 3.5
matrix-multiply 1.0 1.1 1.1
md5 1.0 1.0 1.0
merge 26.9 27.2 27.5
mlyacc 6.0 6.1 6.4
mpuz 2.5 2.5 2.5
nucleic 7.1 5.5 5.4
peek 0.4 0.4 0.4
psdes-random 1.8 1.8 1.8
ratio-regions 5.4 5.6 5.7
ray 1.6 1.6 1.6
raytrace 2.4 2.3 2.3
simple 3.6 3.3 3.4
smith-normal-form 0.4 0.4 0.3
tailfib 5.3 6.5 6.5
tak 4.0 4.3 4.2
tensor 1.6 1.7 1.7
tsp 3.2 2.9 3.0
tyan 8.8 8.6 9.2
vector-concat 1.5 1.6 1.6
vector-rev 2.4 2.4 2.6
vliw 3.2 3.4 3.4
wc-input1 1.2 1.1 1.1
wc-scanStream 1.3 1.3 1.3
zebra 1.0 1.2 1.2
zern 20.9 22.1 21.6
run time ratio
benchmark MLton1 MLton2
barnes-hut 1.0 1.1
boyer 1.1 1.1
checksum 1.0 1.0
count-graphs 1.1 1.1
DLXSimulator 0.9 1.0
fft 0.9 0.9
fib 0.9 0.9
hamlet 1.0 1.1
imp-for 1.1 1.0
knuth-bendix 1.0 1.0
lexgen 0.8 0.8
life 1.4 1.4
logic 1.0 1.1
mandelbrot 1.0 1.0
matrix-multiply 1.2 1.2
md5 1.0 1.0
merge 1.0 1.0
mlyacc 1.0 1.1
mpuz 1.0 1.0
nucleic 0.8 0.8
peek 1.0 1.0
psdes-random 1.0 1.0
ratio-regions 1.0 1.1
ray 1.0 1.0
raytrace 0.9 0.9
simple 0.9 0.9
smith-normal-form 0.9 0.8
tailfib 1.2 1.2
tak 1.1 1.1
tensor 1.1 1.1
tsp 0.9 0.9
tyan 1.0 1.0
vector-concat 1.1 1.1
vector-rev 1.0 1.1
vliw 1.0 1.1
wc-input1 0.9 0.9
wc-scanStream 1.0 1.0
zebra 1.2 1.3
zern 1.1 1.0
size
benchmark MLton0 MLton1 MLton2
barnes-hut 56,943 65,105 65,361
boyer 118,653 128,204 130,412
checksum 22,973 33,484 33,484
count-graphs 44,125 54,284 54,444
DLXSimulator 88,040 99,383 99,511
fft 32,813 42,640 42,640
fib 23,005 33,516 33,516
hamlet 1,104,572 1,217,029 1,218,917
imp-for 23,005 33,548 33,548
knuth-bendix 64,342 76,037 76,069
lexgen 151,749 163,122 163,794
life 39,485 51,916 51,916
logic 80,061 93,308 93,628
mandelbrot 23,037 33,612 33,612
matrix-multiply 23,581 33,996 33,996
md5 32,598 42,629 42,661
merge 24,253 34,828 34,828
mlyacc 470,117 529,346 531,266
mpuz 27,389 37,708 37,708
nucleic 62,013 72,588 72,588
peek 31,542 41,669 41,701
psdes-random 24,349 34,572 34,572
ratio-regions 42,557 52,044 52,236
ray 85,404 94,181 94,629
raytrace 236,937 267,106 267,554
simple 181,965 191,120 191,152
smith-normal-form 137,567 142,821 143,109
tailfib 22,717 33,260 33,260
tak 23,101 33,676 33,676
tensor 56,510 64,828 64,892
tsp 37,878 48,517 49,221
tyan 85,078 96,181 97,301
vector-concat 23,805 33,996 33,996
vector-rev 23,773 33,868 33,868
vliw 299,413 313,070 315,086
wc-input1 48,638 57,698 57,826
wc-scanStream 49,598 58,338 58,530
zebra 109,014 131,901 132,637
zern 30,476 40,175 40,175
-------------------------------------------------------
This sf.net email is sponsored by: OSDN - Tired of that same old
cell phone? Get a new here for FREE!
https://www.inphonic.com/r.asp?r=sourceforge1&refcode1=vs3390
_______________________________________________
MLton-devel mailing list
MLton-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/mlton-devel