performance page
Stephen Weeks
MLton@sourcelight.com
Mon, 8 Oct 2001 10:29:04 -0700
> I just put the new performance page on the web site.
Thanks.
For our own history, here are the numbers.
MLton0 -- mlton 20011006
ML-Kit -- ML-Kit 4.0.0
Moscow-ML -- Moscow-ML 2.0
SML/NJ -- SML/NJ 110.36
run time ratio
benchmark ML-Kit Moscow-ML SML/NJ
barnes-hut 3.9 13.4 0.8
checksum * * 4.3
count-graphs 8.0 26.7 1.7
DLXSimulator 1.9 * *
fft 10.9 35.4 1.2
fib 0.9 4.5 1.0
hamlet * 10.7 1.8
knuth-bendix 4.8 12.6 2.1
lexgen 2.6 6.0 1.7
life 7.2 21.0 1.0
logic * 4.2 0.8
mandelbrot 4.3 31.1 1.3
matrix-multiply 9.4 35.9 3.2
md5 * * 3.3
merge * * 10.6
mlyacc * 7.8 1.7
mpuz 9.5 53.5 3.4
nucleic * 15.0 0.6
peek 6.8 29.9 2.2
psdes-random 17.6 * 4.5
ratio-regions 13.3 48.9 5.6
ray 5.2 13.0 1.0
raytrace * * 2.4
simple 2.5 13.3 1.3
smith-normal-form * * 95.1
tailfib 1.2 21.9 1.4
tak 1.7 7.8 1.4
tensor * * 4.9
tsp 2.6 14.5 1.6
tyan * 13.5 1.0
vector-concat 7.5 13.4 4.7
vector-rev 15.8 45.7 37.3
vliw 2.8 8.7 1.4
wc-input1 13.8 * 8.5
wc-scanStream 15.4 * 3.3
zebra 14.2 28.3 9.1
zern * * 1.7
compile time
benchmark MLton0 ML-Kit Moscow-ML SML/NJ
barnes-hut 2.4 8.1 0.8 1.6
checksum 0.7 * * 0.3
count-graphs 1.8 2.6 0.2 1.1
DLXSimulator 4.1 11.6 * *
fft 1.2 2.1 0.2 1.0
fib 0.7 1.0 0.1 0.2
hamlet 44.2 * 43.2 95.0
knuth-bendix 2.3 5.4 0.4 2.1
lexgen 5.2 10.1 0.8 4.8
life 1.4 2.8 0.2 0.7
logic 5.9 6.9 0.4 1.7
mandelbrot 0.7 1.2 0.1 0.3
matrix-multiply 0.7 1.1 0.1 0.3
md5 1.6 * * 2.1
merge 0.7 1.0 0.1 0.3
mlyacc 20.7 60.8 7.4 23.8
mpuz 1.0 1.4 0.1 0.5
nucleic 3.1 28.4 2.1 2.4
peek 1.1 1.1 0.1 0.3
psdes-random 0.7 1.1 * 0.3
ratio-regions 2.5 4.3 0.4 2.0
ray 3.3 3.7 0.2 1.2
raytrace 8.8 * * 7.0
simple 6.2 14.7 0.9 4.6
smith-normal-form 7.4 * * 3.5
tailfib 0.6 1.0 0.1 0.2
tak 0.6 1.0 0.1 0.2
tensor 2.9 * * 3.5
tsp 1.5 2.7 0.3 0.8
tyan 3.6 * 0.6 2.8
vector-concat 0.7 1.0 0.1 0.2
vector-rev 0.7 1.0 0.1 0.3
vliw 11.8 37.5 3.0 18.9
wc-input1 1.6 1.1 0.1 0.3
wc-scanStream 1.6 1.1 0.1 0.3
zebra 9.2 2.9 0.2 0.8
zern 1.1 * * 0.8
run time
benchmark MLton0 ML-Kit Moscow-ML SML/NJ
barnes-hut 4.9 19.1 66.1 4.0
checksum 4.0 * * 17.1
count-graphs 5.6 44.4 148.4 9.7
DLXSimulator 13.3 25.9 * *
fft 7.1 77.3 250.4 8.2
fib 4.6 4.0 20.9 4.8
hamlet 9.1 * 97.3 16.0
knuth-bendix 8.4 40.8 106.4 17.5
lexgen 12.7 33.0 76.2 21.1
life 10.6 76.3 223.7 10.7
logic 26.2 * 109.5 21.4
mandelbrot 9.1 39.5 282.4 11.7
matrix-multiply 6.7 63.6 242.4 21.7
md5 4.5 * * 15.0
merge 40.0 * * 425.4
mlyacc 10.7 * 83.7 18.0
mpuz 6.3 60.0 336.1 21.6
nucleic 8.4 * 125.3 5.4
peek 4.7 31.6 139.2 10.1
psdes-random 4.6 81.6 * 20.9
ratio-regions 9.0 118.9 438.8 50.4
ray 5.0 26.3 65.5 5.0
raytrace 5.8 * * 13.7
simple 6.8 16.8 90.5 9.2
smith-normal-form 1.1 * * 102.9
tailfib 22.2 25.7 485.4 30.5
tak 10.7 18.3 83.8 15.2
tensor 9.7 * * 47.4
tsp 11.9 30.5 172.2 19.4
tyan 19.8 * 268.2 19.1
vector-concat 7.8 58.8 104.6 36.5
vector-rev 3.2 50.4 146.0 119.2
vliw 6.7 19.0 58.5 9.7
wc-input1 2.8 38.9 * 23.9
wc-scanStream 4.5 68.8 * 14.7
zebra 2.6 37.7 74.9 24.1
zern 37.6 * * 62.3
size
benchmark MLton0 ML-Kit Moscow-ML SML/NJ
barnes-hut 59,793 179,028 94,990 327,672
checksum 20,917 * * 332,504
count-graphs 40,461 109,148 84,575 354,352
DLXSimulator 78,237 178,292 * *
fft 29,441 106,492 84,095 328,712
fib 20,909 67,852 79,878 306,872
hamlet 945,328 * 277,168 1,246,408
knuth-bendix 59,710 114,436 88,439 321,504
lexgen 122,061 226,300 104,883 391,160
life 38,565 99,388 83,390 317,400
logic 147,501 135,228 87,252 330,720
mandelbrot 20,901 100,724 81,341 311,992
matrix-multiply 21,309 117,532 81,879 338,632
md5 34,038 * * 331,792
merge 21,885 68,060 80,091 307,904
mlyacc 409,501 524,460 148,286 684,072
mpuz 26,645 88,732 82,381 304,096
nucleic 60,653 233,124 207,154 351,216
peek 28,542 76,668 81,618 311,016
psdes-random 21,901 83,780 * 313,016
ratio-regions 41,893 110,492 87,485 333,808
ray 66,688 122,500 89,860 382,024
raytrace 159,381 * * 501,864
simple 146,913 193,132 94,397 636,960
smith-normal-form 141,053 * * 483,400
tailfib 20,637 67,740 79,939 306,872
tak 20,957 67,628 79,928 302,776
tensor 62,516 * * 341,024
tsp 33,774 114,844 86,140 319,480
tyan 77,054 * 91,587 367,632
vector-concat 21,557 76,932 80,191 317,128
vector-rev 21,389 77,124 80,073 317,128
vliw 261,417 417,060 135,386 613,456
wc-input1 39,222 143,532 86,900 311,992
wc-scanStream 41,614 143,980 87,076 313,016
zebra 103,502 84,660 83,419 322,536
zern 26,504 * * 335,896