OK - straight comparison with David Mitchell's report:
Processor - AMD A10-7300 Radeon R6, 10 Compute Cores 4C+6G, 1.90 GHz
RAM - 16GB (14.9 GB usable)
64-bit OS, x86 based processor
His 2.5-2.6 GHz seems a lot faster than my 1.9 GHz! Ho hum. But the
avx speed-up is comparable.
ts
6!:2 , 7!:2@]
J805 - no avx
5 ts 'im=: %. rm=: 1000 1000 ?.@$0'
4.440239 69209984
5 ts 'z=: im +/ . * rm'
2.6845591 8391040
J806 - with avx
5 ts 'im=: %. rm=: 1000 1000 ?.@$0'
2.1072457 69211008
5 ts 'z=: im +/ . * rm'
0.62169744 8391040
Sorry, no time for more now
Mike
On 13/03/2017 08:09, David Mitchell wrote:
Processor Intel(R) Core(TM) i7-6500U CPU @ 2.50GHz 2.60 GHz
J806 AVX:
9!:11 ]8
5 timespacex 'im=: %. rm=: 1000 1000 ?.@$0'
0.57594927 60821376
5 timespacex 'z=: im +/ . * rm'
0.16185079 8391040
J805 No AVX:
9!:11 ]8
5 timespacex 'im=: %. rm=: 1000 1000 ?.@$0'
1.0837653 60821376
5 timespacex 'z=: im +/ . * rm'
0.64055261 8391040
On 3/13/2017 01:00, Joey K Tuttle wrote:
Running the same comparisons again, but following Pascal's example
using floating random matrix as well as inverse.
Model Identifier: MacBookPro12,1
Processor Name: Intel Core i5
Processor Speed: 2.7 GHz
Number of Processors: 1
Total Number of Cores: 2
L2 Cache (per Core): 256 KB
L3 Cache: 3 MB
Memory: 8 GB
JVERSION
Engine: j805/j64/darwin
Release: commercial/2016-12-11T08:17:56
Library: 8.05.10
Platform: Darwin 64
Installer: J805 install
InstallPath: /applications/j64-805
Contact: www.jsoftware.com <http://www.jsoftware.com/>
timex
6!:2 , 7!:2@]
5 timex 'im=: %. rm=: 1000 1000 ?.@$0'
1.6871406 60821376
5 timex 'z=: im +/ . * rm'
0.9249066 8391040
4 4 {. z
1.000000049 1.166244801e_8 3.450154074e_8 3.349510003e_8
_1.510177019e_7 0.999999964 _1.066236264e_7 _1.035110442e_7
_1.357366599e_7 _3.239824409e_8 0.9999999042 _9.304124804e_8
_5.419372251e_8 _1.29358585e_8 _3.826158634e_8 0.9999999629
~~~~~~
JVERSION
Engine: j806/j64avx/darwin
Beta-1: commercial/2017-03-09T11:59:57
Library: 8.06.01
Platform: Darwin 64
Installer: J806 install
InstallPath: /applications/j64-806
Contact: www.jsoftware.com <http://www.jsoftware.com/>
5 timex 'im=: %. rm=: 1000 1000 ?.@$0'
0.750128 60821376
5 timex 'z=: im +/ . * rm'
0.229624 8391040
4 4 {. z
1.000000022 _5.609759768e_9 1.444602749e_8 1.202453515e_8
_6.64858355e_8 1.000000017 _4.463911907e_8 _3.715585528e_8
_5.975335071e_8 1.558899854e_8 0.9999999599 _3.339379262e_8
_2.38620288e_8 6.217513615e_9 _1.602165334e_8 0.9999999867
Speed differences stay roughly the same - results are different from
each other as well as from my first example where the random matrix
was integer.
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
