First test of my neural net app is much slower - 20 sec vs. 8 sec for 805 stable 😩. ????
Sent from my iPhone > On Apr 17, 2017, at 1:38 PM, David Mitchell <[email protected]> wrote: > > Here are my results with beta-3: > > 2017 4 16 4 41 > Intel(R) Core(TM) i7-6500U CPU @ 2.50GHz > j805/j64/windows/beta-12/commercial/www.jsoftware.com/2016-09-30T15:50:47 > j806/j64avx/windows/beta-3/commercial/www.jsoftware.com/2017-04-10T18:03:23 > intsr (small range) special code avoids hash - intbr (big range) > float0 tests use !.0 where appropriate > N in tables below indicate avx JE runs N times faster than 805 > > 'type' set 1e7 1e3 > intsr intbr char float float0 test > 1.4 1.9 3.4 2.4 3.2 a i. a > 19.2 12.7 24.5 2.0 23.9 a i. b > 2.1 4.3 7.0 7.0 6.3 b i. a > 3.0 5.5 8.2 8.3 8.7 a e. b > 10.2 12.6 25.5 1.9 23.4 b e. a > 4.0 6.2 9.2 7.7 9.7 a (+/@:e.) b > 1.1 6.2 8.8 31.1 8.7 a (e. i. 1:) b > 2.2 1.8 2.8 2.1 1.9 ~.a > 2.8 1.9 3.8 2.5 2.3 ~:a > 1.1 1.1 1.3 1.1 1.1 /:a > 1.8 1.1 1.5 1.1 1.0 /:~a > > 'type' set 1e5 1e3 > intsr intbr char float float0 test > 2.5 3.2 4.7 4.2 3.1 a i. a > 7.6 5.9 10.5 4.0 6.6 a i. b > 2.2 4.2 6.7 6.8 6.3 b i. a > 2.6 5.3 7.7 8.0 7.9 a e. b > 3.5 5.9 10.0 5.2 8.1 b e. a > 4.3 5.9 8.5 8.5 9.2 a (+/@:e.) b > 1.6 3.4 7.9 18.3 8.2 a (e. i. 1:) b > 1.3 3.3 2.8 3.6 2.6 ~.a > 1.5 3.7 5.6 4.9 3.7 ~:a > 1.5 1.3 1.3 1.3 1.2 /:a > 3.4 1.3 1.0 1.2 1.1 /:~a > > 'type' set '1e3' > mmint mmfloat mmcomplex test > 4.7 3.7 3.2 a +/ . * b > > === > rep'' > 2017 4 16 11 42 > Intel(R) Core(TM) i7-6500U CPU @ 2.50GHz > j805/j64/linux/release/commercial/www.jsoftware.com/2016-12-11T08:02:52 > j806/j64avx/linux/beta-3/commercial/www.jsoftware.com/2017-04-10T17:51:14 > intsr (small range) special code avoids hash - intbr (big range) > float0 tests use !.0 where appropriate > N in tables below indicate avx JE runs N times faster than 805 > > 'type' set 1e7 1e3 > intsr intbr char float float0 test > 1.6 1.8 3.3 2.1 3.3 a i. a > 20.6 12.3 26.1 2.0 22.4 a i. b > 3.3 6.7 7.7 7.3 10.7 b i. a > 4.9 7.8 8.3 7.3 13.5 a e. b > 13.9 12.1 28.7 2.0 24.3 b e. a > 4.8 8.0 8.6 6.9 12.6 a (+/@:e.) b > 4.3 6.8 9.0 59.9 13.1 a (e. i. 1:) b > 2.9 1.9 2.9 1.6 1.5 ~.a > 3.5 2.0 3.4 2.0 2.0 ~:a > 1.0 1.0 1.2 1.1 1.1 /:a > 1.2 1.2 2.8 1.1 1.1 /:~a > > 'type' set 1e5 1e3 > intsr intbr char float float0 test > 1.9 4.1 5.6 2.8 5.0 a i. a > 4.5 5.6 15.5 3.3 7.6 a i. b > 5.8 8.7 8.9 6.3 10.6 b i. a > 4.0 8.2 8.9 6.7 13.4 a e. b > 3.4 5.2 15.2 3.3 8.1 b e. a > 5.5 8.9 9.7 6.5 12.9 a (+/@:e.) b > 1.5 4.0 7.0 21.3 13.3 a (e. i. 1:) b > 1.3 3.5 6.4 2.4 2.0 ~.a > 1.2 3.9 4.6 2.9 2.9 ~:a > 0.9 1.3 1.3 1.2 1.2 /:a > 1.6 1.4 1.2 1.2 1.4 /:~a > > 'type' set '1e3' > mmint mmfloat mmcomplex test > > 23.2 4.8 3.3 a +/ . * b > >> On 4/11/2017 15:18, Eric Iverson wrote: >> 806 beta-3 available. >> >> Comments from original announcement are repeated here for emphasis. >> >> 806 will be primarily a performance release. This is the first J release >> where hardware features are directly used for performance. Previous >> releases depended on excellent code and smart algorithms. With Advanced >> Vector Extensions (AVX) Intel finally (first hardware released in 2011) has >> hardware that seems to have J, at least partly, in mind. >> >> A rough benchmark report is at the end of this message. It has been a long >> time since we've been able to brag of a factor of 10 speedup in a primitive. >> >> Improvements in i. and related areas are important in J, but faster >> crunching is usually overwhelmed by all the housekeeping in an application. >> Some things run 10 times faster, but your application won't. >> >> Please get involved in the beta program, it helps make a better product for >> everyone. >> >> And give big thanks to Henry Rich for this core JE development! >> >> *** >> Follow web site download links. There have been changes. Please follow the >> directions and report any problems. >> >> These releases are only for windows/osx/linux intel/amd 64 and include only >> an avx binary. >> >> The J Engine load will fail if the hardware/OS does not support avx. >> >> *** benchmark report - ~addons/ide/jhs/misc/avx.ijs >> >> 2017 4 11 15 1 >> Intel(R) Core(TM) i7-7700 CPU @ 3.60GHz >> j805/j64/linux/release-a/commercial/www.jsoftware.com/2017-02-26T16:47:20 >> j806/j64avx/linux/beta-3/commercial/www.jsoftware.com/2017-04-10T17:51:14 >> intsr (small range) special code avoids hash - intbr (big range) >> float0 tests use !.0 where appropriate >> N in tables below indicate avx JE runs N times faster than 805 >> >> 'type' set 1e7 1e3 >> intsr intbr char float float0 test >> 1.3 2.0 4.1 2.1 3.6 a i. a >> 12.9 10.7 25.8 2.1 20.5 a i. b >> 3.4 7.3 8.6 6.7 10.6 b i. a >> 5.3 8.2 9.1 7.2 13.2 a e. b >> 6.4 10.6 25.9 2.1 20.4 b e. a >> 5.3 8.8 9.5 7.0 13.2 a (+/@:e.) b >> 4.4 6.5 9.6 53.4 13.2 a (e. i. 1:) b >> 1.7 1.9 3.8 1.7 1.7 ~.a >> 1.6 2.1 4.1 2.2 2.0 ~:a >> 1.1 1.3 1.3 1.2 1.2 /:a >> 1.3 2.1 2.2 1.9 1.9 /:~a >> >> 'type' set 1e5 1e3 >> intsr intbr char float float0 test >> 2.0 3.5 5.3 2.4 4.7 a i. a >> 4.2 4.7 9.2 3.2 6.9 a i. b >> 5.4 8.0 8.9 5.2 11.7 b i. a >> 5.8 7.7 9.3 6.7 12.7 a e. b >> 4.1 4.7 9.2 3.2 6.9 b e. a >> 5.7 8.3 9.7 6.5 12.5 a (+/@:e.) b >> 2.0 4.0 7.8 21.0 12.6 a (e. i. 1:) b >> 1.5 3.3 4.5 1.9 1.9 ~.a >> 1.3 3.4 5.2 2.9 2.9 ~:a >> 2.1 2.1 1.3 1.9 1.9 /:a >> 1.7 2.3 1.4 2.1 2.1 /:~a >> >> 'type' set '1e3' >> mmint mmfloat mmcomplex test >> 27.6 23.7 17.9 a +/ . * b >> ---------------------------------------------------------------------- >> For information about J forums see http://www.jsoftware.com/forums.htm >> > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
