Board index » delphi » Fastcode CompareStr B&V 2.2

Fastcode CompareStr B&V 2.2


2005-10-20 02:17:40 AM
delphi32
Hi
Embedded version info is wrong.
Opteron 240 benchmark results.
Best regards
Dennis
CompareStr_PLR_IA32_14_c 4 464 583 525 1572
CompareStr_PLR_IA32_14_b 0 468 592 519 1579
CompareStr_PLR_IA32_12_a 0 466 604 520 1590
CompareStr_PLR_IA32_12_d C 464 614 513 1591
CompareStr_PLR_IA32_15_a 8 513 576 542 1631
CompareStr_JOH_IA32_1_b 8 505 589 538 1632
CompareStr_PLR_IA32_16_a C 515 593 541 1649
CompareStr_PLR_IA32_15_d 4 513 606 541 1660
CompareStr_PLR_IA32_14_a C 466 697 519 1682
CompareStr_PLR_IA32_12_c 8 469 732 520 1721
CompareStr_JOH_IA32_1_c 4 510 696 533 1739
CompareStr_PLR_IA32_15_b 0 512 699 540 1751
CompareStr_PLR_IA32_16_d 8 515 729 545 1789
CompareStr_PLR_IA32_14_d 8 472 841 525 1838
CompareStr_PLR_IA32_12_b 4 466 869 523 1858
CompareStr_PLR_IA32_16_b 4 518 840 540 1898
CompareStr_JOH_IA32_1_a 0 513 848 540 1901
CompareStr_PLR_IA32_15_c C 514 857 537 1908
CompareStr_JOH_IA32_1_d C 512 860 540 1912
CompareStr_PLR_IA32_16_c 0 517 859 543 1919
CompareStr_PLR_PAS_1_d 0 768 1053 852 2673
CompareStr_PLR_PAS_1_b 0 777 1067 851 2695
CompareStr_PLR_PAS_1_c 0 777 1066 861 2704
CompareStr_PLR_PAS_1_a 0 777 1066 861 2704
CompareStr_IA32_EWC_3_d 8 1001 648 1216 2865
CompareStr_IA32_EWC_3_b 0 1016 645 1210 2871
CompareStr_IA32_EWC_3_a 4 1036 656 1277 2969
CompareStr_IA32_EWC_3_c C 966 906 1222 3094
CompareStr_DKC_MMX_3_c 0 1142 673 1374 3189
CompareStr_DKC_SSE2_3_d 8 1226 740 1281 3247
CompareStr_DKC_SSE2_3_a C 1235 740 1284 3259
CompareStr_DKC_SSE2_3_b 0 1265 723 1309 3297
CompareStr_DKC_MMX_5_a 8 1179 767 1360 3306
CompareStr_DKC_MMX_4_b 0 1167 783 1375 3325
CompareStr_DKC_MMX_3_a 8 1182 799 1424 3405
CompareStr_DKC_MMX_5_c C 1223 795 1408 3426
CompareStr_DKC_MMX_5_b 0 1250 744 1433 3427
CompareStr_DKC_MMX_3_d C 1247 719 1481 3447
CompareStr_DKC_MMX_4_a C 1254 712 1492 3458
CompareStr_DKC_MMX_4_d 8 1248 754 1470 3472
CompareStr_DKC_IA32_20_a C 1229 756 1490 3475
CompareStr_DKC_SSE2_3_c 4 1349 737 1457 3543
CompareStr_DKC_MMX_3_b 4 1256 790 1508 3554
CompareStr_DKC_MMX_5_d 4 1321 728 1521 3570
CompareStr_DKC_MMX_4_c 4 1284 806 1506 3596
CompareStr_DKC_IA32_20_b 8 1326 819 1589 3734
CompareStr_DKC_Pas_11_c C 1422 825 1556 3803
CompareStr_DKC_Pas_11_a 8 1392 881 1533 3806
CompareStr_DKC_Pas_11_d 4 1423 822 1602 3847
CompareStr_DKC_Pas_12_c 4 1434 833 1662 3929
CompareStr_DKC_Pas_11_b 0 1457 865 1649 3971
CompareStr_DKC_IA32_20_d 0 1507 730 1883 4120
CompareStr_DKC_Pas_12_d 0 1543 911 1741 4195
CompareStr_LBG_IA32ext_1_c C 1335 1386 1528 4249
CompareStr_DKC_Pas_12_b 8 1616 817 1844 4277
CompareStr_DKC_Pas_12_a C 1580 906 1804 4290
CompareStr_LBG_IA32ext_1_b 8 1349 1403 1547 4299
CompareStr_LBG_IA32ext_1_a 4 1352 1406 1549 4307
CompareStr_LBG_IA32ext_1_d 0 1384 1402 1569 4355
CompareStrRTL 8 1338 1383 1673 4394
CompareStr_DKC_IA32_20_c 4 1614 689 2100 4403
CompareStr_PAS_EWC_2_d 0 1589 1101 1942 4632
CompareStr_PAS_EWC_2_c C 1571 1183 1937 4691
CompareStr_PAS_EWC_2_a 4 1626 1138 1967 4731
CompareStr_PAS_EWC_2_b 8 1612 1175 1983 4770
CompareStr_DKC_SSE3_3_d X 9999 9999 9999 9999
CompareStr_DKC_SSE3_3_c X 9999 9999 9999 9999
CompareStr_DKC_SSE3_3_b X 9999 9999 9999 9999
CompareStr_DKC_SSE3_3_a X 9999 9999 9999 9999
 
 

Re:Fastcode CompareStr B&V 2.2

Hi
Todo
Correct B&V and make new release.
Release at Fastcode site.
Run validation and benchmarks
Update spreadsheet
Release at Fastcode site.
Who is doing what and when?
Best regards
Dennis
 

Re:Fastcode CompareStr B&V 2.2

"Dennis" <XXXX@XXXXX.COM>writes
Quote
Todo

Correct B&V and make new release.
Don't know who does that.
Quote
Release at Fastcode site.
That's me! Just let me know when its ready.
Quote
Run validation and benchmarks
Not me.
Quote
Update spreadsheet
Not me.
Quote
Release at Fastcode site.
That's me! Just let me know when ready...
 

Re:Fastcode CompareStr B&V 2.2

Hi All
There is work that needs to be done.
Quote
>Correct B&V and make new release.
>Run validation and benchmarks
>Update spreadsheet
I see this release as a test of the B&V release cycle for future Fastcode
work. To me it looks like nobody is interested in doing what is necessary
to
keep the project on the track.
The algorithm for the release of a B&V does not have to be settled at this
early stage of Fastcode II, but I'd like to see some progress. As I am
not the manager anymore somebody needs to step forward make something
happen.
I could suggest this algorithm:
One person is responsible for doing release X.Y of B&V Z.
He makes the B&V changes.
He releases the B&V in attachments.
He sends the B&V to the webmanager
He collects benchmark results for all targets
He updates the spreadsheet
He sends it to the webmanager
He updates the winnerslist
He sends it to the webmanager
The webmanager releases the B&V
The webmanager releases the resultsspreadsheet
The webmanager releases the winnerslist.
I think that it is important to get the results spreadsheet and
winnerslist
released very soon after the B&V release. In Fastcode I it was done at the
same time.
Regards
Dennis
----------------------------------------
Jeg beskyttes af den gratis SPAMfighter til privatbrugere.
Den har indtil videre sparet mig for at få 1555 spam-mails.
Betalende brugere får ikke denne besked i deres e-mails.
Hent gratis SPAMfighter her: www.spamfighter.dk
 

Re:Fastcode CompareStr B&V 2.2

Hi Dennis
Dennis writes:
Quote

Todo

Correct B&V
That is it specific there is to be Corrected ?
Best regards
Lars G