Windows XP 32bit q9400 2.6GHz Release build (no PG optimizations). Test runs in background, ccbench modified to run both bz2 and sha1. bfs.patch Results using all (4) cores: ---------------------------- F:\dev>f:\build\python\bfs\PCbuild\python.exe y:\ccbench.py == CPython 3.2a0.0 (py3k) == == x86 Windows on 'x86 Family 6 Model 23 Stepping 10, GenuineIntel' == --- Throughput --- Pi calculation (Python) threads= 1: 840 iterations/s. balance threads= 2: 843 (100%) 0.0166 threads= 3: 837 ( 99%) 0.0204 threads= 4: 830 ( 98%) 0.0235 regular expression (C) threads= 1: 480 iterations/s. balance threads= 2: 476 ( 99%) 0.0000 threads= 3: 470 ( 97%) 0.0369 threads= 4: 456 ( 95%) 0.0502 bz2 compression (C) threads= 1: 510 iterations/s. balance threads= 2: 1010 (198%) 0.0140 threads= 3: 1498 (293%) 0.0115 threads= 4: 1970 (386%) 0.0102 SHA1 hashing (C) threads= 1: 2390 iterations/s. balance threads= 2: 4702 (196%) 0.0068 threads= 3: 7012 (293%) 0.0056 threads= 4: 9347 (391%) 0.0048 --- Latency --- Background CPU task: Pi calculation (Python) CPU threads=0: 0 ms. (std dev: 0 ms.) CPU threads=1: 0 ms. (std dev: 0 ms.) CPU threads=2: 0 ms. (std dev: 0 ms.) CPU threads=3: 0 ms. (std dev: 3 ms.) CPU threads=4: 0 ms. (std dev: 0 ms.) Background CPU task: regular expression (C) CPU threads=0: 0 ms. (std dev: 0 ms.) CPU threads=1: 0 ms. (std dev: 0 ms.) CPU threads=2: 0 ms. (std dev: 0 ms.) CPU threads=3: 0 ms. (std dev: 0 ms.) CPU threads=4: 0 ms. (std dev: 0 ms.) Background CPU task: bz2 compression (C) CPU threads=0: 0 ms. (std dev: 0 ms.) CPU threads=1: 0 ms. (std dev: 0 ms.) CPU threads=2: 0 ms. (std dev: 0 ms.) CPU threads=3: 0 ms. (std dev: 0 ms.) CPU threads=4: 3 ms. (std dev: 8 ms.) Background CPU task: SHA1 hashing (C) CPU threads=0: 0 ms. (std dev: 0 ms.) CPU threads=1: 0 ms. (std dev: 0 ms.) CPU threads=2: 0 ms. (std dev: 0 ms.) CPU threads=3: 0 ms. (std dev: 0 ms.) CPU threads=4: 0 ms. (std dev: 0 ms.) --- I/O bandwidth --- Background CPU task: Pi calculation (Python) CPU threads=0: 8228.6 packets/s. CPU threads=1: 5413.9 ( 65 %) CPU threads=2: 4640.7 ( 56 %) CPU threads=3: 3460.7 ( 42 %) CPU threads=4: 3400.8 ( 41 %) Results using single core: -------------------------- F:\dev>f:\build\python\bfs\PCbuild\python.exe y:\ccbench.py == CPython 3.2a0.0 (py3k) == == x86 Windows on 'x86 Family 6 Model 23 Stepping 10, GenuineIntel' == --- Throughput --- Pi calculation (Python) threads= 1: 830 iterations/s. balance threads= 2: 840 (101%) 0.0337 threads= 3: 833 (100%) 0.0000 threads= 4: 816 ( 98%) 0.0241 regular expression (C) threads= 1: 476 iterations/s. balance threads= 2: 472 ( 99%) 0.0000 threads= 3: 470 ( 98%) 0.0369 threads= 4: 456 ( 95%) 0.0502 bz2 compression (C) threads= 1: 510 iterations/s. balance threads= 2: 500 ( 98%) 0.0000 threads= 3: 506 ( 99%) 0.0000 threads= 4: 492 ( 96%) 0.0462 SHA1 hashing (C) threads= 1: 2367 iterations/s. balance threads= 2: 2350 ( 99%) 0.0060 threads= 3: 2375 (100%) 0.2443 threads= 4: 2360 ( 99%) 0.0138 --- Latency --- Background CPU task: Pi calculation (Python) CPU threads=0: 0 ms. (std dev: 0 ms.) CPU threads=1: 0 ms. (std dev: 0 ms.) CPU threads=2: 0 ms. (std dev: 0 ms.) CPU threads=3: 0 ms. (std dev: 0 ms.) CPU threads=4: 0 ms. (std dev: 0 ms.) Background CPU task: regular expression (C) CPU threads=0: 0 ms. (std dev: 0 ms.) CPU threads=1: 0 ms. (std dev: 0 ms.) CPU threads=2: 0 ms. (std dev: 0 ms.) CPU threads=3: 0 ms. (std dev: 0 ms.) CPU threads=4: 0 ms. (std dev: 0 ms.) Background CPU task: bz2 compression (C) CPU threads=0: 0 ms. (std dev: 0 ms.) CPU threads=1: 0 ms. (std dev: 0 ms.) CPU threads=2: 0 ms. (std dev: 0 ms.) CPU threads=3: 0 ms. (std dev: 0 ms.) CPU threads=4: 0 ms. (std dev: 0 ms.) Background CPU task: SHA1 hashing (C) CPU threads=0: 0 ms. (std dev: 0 ms.) CPU threads=1: 0 ms. (std dev: 0 ms.) CPU threads=2: 0 ms. (std dev: 0 ms.) CPU threads=3: 0 ms. (std dev: 0 ms.) CPU threads=4: 0 ms. (std dev: 0 ms.) --- I/O bandwidth --- Background CPU task: Pi calculation (Python) CPU threads=0: 9461.7 packets/s. CPU threads=1: 6707.6 ( 70 %) CPU threads=2: 6630.3 ( 70 %) CPU threads=3: 6936.4 ( 73 %) CPU threads=4: 4861.7 ( 51 %) ============================================================= gilinter2.patch single core: high latency, low IO bandwidth. Results using all (4) cores: ---------------------------- F:\dev>f:\build\python\gilinter\PCbuild\python.exe y:\ccbench.py == CPython 3.2a0.0 (py3k) == == x86 Windows on 'x86 Family 6 Model 23 Stepping 10, GenuineIntel' == --- Throughput --- Pi calculation (Python) threads= 1: 850 iterations/s. balance threads= 2: 850 (100%) 0.0832 threads= 3: 846 ( 99%) 0.0533 threads= 4: 850 (100%) 0.1352 regular expression (C) threads= 1: 452 iterations/s. balance threads= 2: 452 (100%) 0.0615 threads= 3: 452 (100%) 0.3012 threads= 4: 443 ( 97%) 0.3025 bz2 compression (C) threads= 1: 512 iterations/s. balance threads= 2: 1010 (197%) 0.0140 threads= 3: 1480 (289%) 0.0117 threads= 4: 1920 (374%) 0.0241 SHA1 hashing (C) threads= 1: 2371 iterations/s. balance threads= 2: 4745 (200%) 0.0089 threads= 3: 7027 (296%) 0.0067 threads= 4: 9295 (392%) 0.0097 --- Latency --- Background CPU task: Pi calculation (Python) CPU threads=0: 0 ms. (std dev: 0 ms.) CPU threads=1: 0 ms. (std dev: 3 ms.) CPU threads=2: 0 ms. (std dev: 0 ms.) CPU threads=3: 0 ms. (std dev: 0 ms.) CPU threads=4: 0 ms. (std dev: 0 ms.) Background CPU task: regular expression (C) CPU threads=0: 0 ms. (std dev: 0 ms.) CPU threads=1: 0 ms. (std dev: 3 ms.) CPU threads=2: 0 ms. (std dev: 3 ms.) CPU threads=3: 0 ms. (std dev: 0 ms.) CPU threads=4: 0 ms. (std dev: 0 ms.) Background CPU task: bz2 compression (C) CPU threads=0: 0 ms. (std dev: 0 ms.) CPU threads=1: 0 ms. (std dev: 0 ms.) CPU threads=2: 0 ms. (std dev: 0 ms.) CPU threads=3: 0 ms. (std dev: 0 ms.) CPU threads=4: 0 ms. (std dev: 0 ms.) Background CPU task: SHA1 hashing (C) CPU threads=0: 0 ms. (std dev: 0 ms.) CPU threads=1: 0 ms. (std dev: 0 ms.) CPU threads=2: 0 ms. (std dev: 0 ms.) CPU threads=3: 0 ms. (std dev: 0 ms.) CPU threads=4: 0 ms. (std dev: 0 ms.) --- I/O bandwidth --- Background CPU task: Pi calculation (Python) CPU threads=0: 7473.2 packets/s. CPU threads=1: 4794.1 ( 64 %) CPU threads=2: 3804.7 ( 50 %) CPU threads=3: 2858.9 ( 38 %) CPU threads=4: 2556.8 ( 34 %) Results using single Core: -------------------------- F:\dev>f:\build\python\gilinter\PCbuild\python.exe y:\ccbench.py == CPython 3.2a0.0 (py3k) == == x86 Windows on 'x86 Family 6 Model 23 Stepping 10, GenuineIntel' == --- Throughput --- Pi calculation (Python) threads= 1: 840 iterations/s. balance threads= 2: 850 (101%) 0.0166 threads= 3: 860 (102%) 0.0201 threads= 4: 833 ( 99%) 0.1739 regular expression (C) threads= 1: 452 iterations/s. balance threads= 2: 456 (100%) 0.1230 threads= 3: 450 ( 99%) 0.0667 threads= 4: 449 ( 99%) 0.1506 bz2 compression (C) threads= 1: 505 iterations/s. balance threads= 2: 502 ( 99%) 0.0277 threads= 3: 500 ( 98%) 0.0346 threads= 4: 488 ( 96%) 0.1386 SHA1 hashing (C) threads= 1: 2352 iterations/s. balance threads= 2: 2350 ( 99%) 0.0181 threads= 3: 2330 ( 99%) 0.0743 threads= 4: 2340 ( 99%) 0.1298 --- Latency --- Background CPU task: Pi calculation (Python) CPU threads=0: 0 ms. (std dev: 0 ms.) CPU threads=1: 14 ms. (std dev: 4 ms.) CPU threads=2: 42 ms. (std dev: 32 ms.) CPU threads=3: 134 ms. (std dev: 115 ms.) CPU threads=4: 147 ms. (std dev: 172 ms.) Background CPU task: regular expression (C) CPU threads=0: 0 ms. (std dev: 0 ms.) CPU threads=1: 14 ms. (std dev: 4 ms.) CPU threads=2: 33 ms. (std dev: 26 ms.) CPU threads=3: 28 ms. (std dev: 49 ms.) CPU threads=4: 174 ms. (std dev: 198 ms.) Background CPU task: bz2 compression (C) CPU threads=0: 0 ms. (std dev: 0 ms.) CPU threads=1: 0 ms. (std dev: 0 ms.) CPU threads=2: 1 ms. (std dev: 4 ms.) CPU threads=3: 8 ms. (std dev: 27 ms.) CPU threads=4: 7 ms. (std dev: 13 ms.) Background CPU task: SHA1 hashing (C) CPU threads=0: 0 ms. (std dev: 0 ms.) CPU threads=1: 0 ms. (std dev: 0 ms.) CPU threads=2: 3 ms. (std dev: 7 ms.) CPU threads=3: 3 ms. (std dev: 10 ms.) CPU threads=4: 7 ms. (std dev: 15 ms.) --- I/O bandwidth --- Background CPU task: Pi calculation (Python) CPU threads=0: 9068.4 packets/s. CPU threads=1: 52.2 ( 0 %) CPU threads=2: 22.4 ( 0 %) CPU threads=3: 0.8 ( 0 %) CPU threads=4: 4.9 ( 0 %) ================================================================ dabeaz_gil.patch single core: low IO bandwidth. 4 cores: throughput threads starvation (balance), some latency, low IO bandwidth. Results using all (4) cores: ---------------------------- F:\dev>f:\build\python\dabeaz\PCbuild\python.exe y:\ccbench.py == CPython 3.2a0.0 (py3k) == == x86 Windows on 'x86 Family 6 Model 23 Stepping 10, GenuineIntel' == --- Throughput --- Pi calculation (Python) threads= 1: 803 iterations/s. balance threads= 2: 803 ( 99%) 0.0873 threads= 3: 800 ( 99%) 0.1732 threads= 4: 780 ( 97%) 0.0296 regular expression (C) threads= 1: 456 iterations/s. balance threads= 2: 460 (100%) 0.3074 threads= 3: 449 ( 98%) 0.3765 threads= 4: 436 ( 95%) 0.7385 bz2 compression (C) threads= 1: 515 iterations/s. balance threads= 2: 1010 (195%) 0.0140 threads= 3: 1486 (288%) 0.0115 threads= 4: 1970 (381%) 0.0102 SHA1 hashing (C) threads= 1: 2361 iterations/s. balance threads= 2: 4715 (199%) 0.0000 threads= 3: 7005 (296%) 0.0000 threads= 4: 9230 (390%) 0.0000 --- Latency --- Background CPU task: Pi calculation (Python) CPU threads=0: 0 ms. (std dev: 0 ms.) CPU threads=1: 2 ms. (std dev: 5 ms.) CPU threads=2: 3 ms. (std dev: 9 ms.) CPU threads=3: 3 ms. (std dev: 8 ms.) CPU threads=4: 3 ms. (std dev: 7 ms.) Background CPU task: regular expression (C) CPU threads=0: 0 ms. (std dev: 0 ms.) CPU threads=1: 8 ms. (std dev: 20 ms.) CPU threads=2: 10 ms. (std dev: 21 ms.) CPU threads=3: 5 ms. (std dev: 14 ms.) CPU threads=4: 10 ms. (std dev: 22 ms.) Background CPU task: bz2 compression (C) CPU threads=0: 0 ms. (std dev: 0 ms.) CPU threads=1: 0 ms. (std dev: 0 ms.) CPU threads=2: 0 ms. (std dev: 0 ms.) CPU threads=3: 0 ms. (std dev: 0 ms.) CPU threads=4: 0 ms. (std dev: 0 ms.) Background CPU task: SHA1 hashing (C) CPU threads=0: 0 ms. (std dev: 0 ms.) CPU threads=1: 0 ms. (std dev: 0 ms.) CPU threads=2: 0 ms. (std dev: 0 ms.) CPU threads=3: 0 ms. (std dev: 0 ms.) CPU threads=4: 0 ms. (std dev: 0 ms.) --- I/O bandwidth --- Background CPU task: Pi calculation (Python) CPU threads=0: 7801.8 packets/s. CPU threads=1: 59.7 ( 0 %) CPU threads=2: 595.7 ( 7 %) CPU threads=3: 72.1 ( 0 %) CPU threads=4: 94.5 ( 1 %) Results using single core: -------------------------- F:\dev>f:\build\python\dabeaz\PCbuild\python.exe y:\ccbench.py == CPython 3.2a0.0 (py3k) == == x86 Windows on 'x86 Family 6 Model 23 Stepping 10, GenuineIntel' == --- Throughput --- Pi calculation (Python) threads= 1: 810 iterations/s. balance threads= 2: 803 ( 99%) 0.0524 threads= 3: 790 ( 97%) 0.0219 threads= 4: 800 ( 98%) 0.0408 regular expression (C) threads= 1: 452 iterations/s. balance threads= 2: 456 (100%) 0.0000 threads= 3: 456 (100%) 0.0377 threads= 4: 446 ( 98%) 0.0444 bz2 compression (C) threads= 1: 510 iterations/s. balance threads= 2: 502 ( 98%) 0.0277 threads= 3: 496 ( 97%) 0.0346 threads= 4: 480 ( 94%) 0.0962 SHA1 hashing (C) threads= 1: 2337 iterations/s. balance threads= 2: 2360 (100%) 0.0120 threads= 3: 2340 (100%) 0.0128 threads= 4: 2340 (100%) 0.0171 --- Latency --- Background CPU task: Pi calculation (Python) CPU threads=0: 0 ms. (std dev: 0 ms.) CPU threads=1: 0 ms. (std dev: 3 ms.) CPU threads=2: 0 ms. (std dev: 0 ms.) CPU threads=3: 0 ms. (std dev: 0 ms.) CPU threads=4: 0 ms. (std dev: 0 ms.) Background CPU task: regular expression (C) CPU threads=0: 0 ms. (std dev: 0 ms.) CPU threads=1: 0 ms. (std dev: 3 ms.) CPU threads=2: 0 ms. (std dev: 0 ms.) CPU threads=3: 0 ms. (std dev: 0 ms.) CPU threads=4: 0 ms. (std dev: 0 ms.) Background CPU task: bz2 compression (C) CPU threads=0: 0 ms. (std dev: 0 ms.) CPU threads=1: 0 ms. (std dev: 0 ms.) CPU threads=2: 0 ms. (std dev: 0 ms.) CPU threads=3: 2 ms. (std dev: 5 ms.) CPU threads=4: 3 ms. (std dev: 6 ms.) Background CPU task: SHA1 hashing (C) CPU threads=0: 0 ms. (std dev: 0 ms.) CPU threads=1: 0 ms. (std dev: 0 ms.) CPU threads=2: 0 ms. (std dev: 0 ms.) CPU threads=3: 0 ms. (std dev: 0 ms.) CPU threads=4: 0 ms. (std dev: 0 ms.) --- I/O bandwidth --- Background CPU task: Pi calculation (Python) CPU threads=0: 8496.4 packets/s. CPU threads=1: 2151.8 ( 25 %) CPU threads=2: 31.5 ( 0 %) CPU threads=3: 16.5 ( 0 %) CPU threads=4: 22.9 ( 0 %)