dalmem throughput tests
These tests are performed for most dalmem functions, and best results for fastest variant are displayed here.
Currently (Version 1.0.1) 32KB buffer is used, so results are for cached sets (RAM speed shouldn't affect these tests results).
Every function block (10 consecutive calls) is called 50 times, and best time is accounted. 50x10 calls usually ensure the CPU is finally running at its fastest speed.
If you want to reproduce these tests on your own machines, please download dalib source code (tested only on unix), compile and test yourself (see README and INSTALL files).
Raw test results:
Test environment: 32.0KB buffer, every test is performed 50 times with 10 iterations, seed is just half the buffer. Version 1.0.1
+--------- Option0 ---------+--------- Option1 ---------+--------- Option2 ---------+--------- Option3 ---------+Best
Function name |F us B/s S us B/s|F us B/s S us B/s|F us B/s S us B/s|F us B/s S us B/s| F S
memsearch8 R| 18.9 1.61G 9.4 1.62G| 18.9 1.61G 9.4 1.62G| 18.9 1.61G 9.4 1.62G| 1.7 17.9G 0.9 16.9G| 3 3
memsearch16 R| 9.4 3.24G 4.7 3.24G| 9.4 3.24G 4.7 3.24G| 9.4 3.24G 4.7 3.24G| 1.7 17.9G 0.9 16.9G| 3 3
memsearch32 R| 4.7 6.49G 2.3 6.63G| 4.7 6.49G 2.3 6.63G| 4.7 6.49G 2.3 6.63G| 1.7 17.9G 0.9 16.9G| 3 3
memsearch64 R| 2.3 13.2G 1.1 13.8G| | | | 0 0
memsearch8r R| 35.7 854M 17.4 876M| 2.4 12.7G 1.2 12.7G| | | 1 1
memsearch16r R| 11.1 2.74G 5.5 2.77G| | | | 0 0
memsearch32r R| 6.0 5.08G 3.1 4.92G| | | | 0 0
memsearch64r R| 3.6 8.47G 1.9 8.03G| | | | 0 0
memsearch8ri R| 28.0 1.08G 13.5 1.13G| 2.3 13.2G 1.2 12.7G| | | 1 1
memsearch16ri R| 11.0 2.77G 5.4 2.82G| | | | 0 0
memsearch32ri R| 6.1 5.00G 3.1 4.92G| | | | 0 0
memsearch64ri R| 3.7 8.24G 1.9 8.03G| | | | 0 0
bitmapsearch8 R| 28.4 1.07G 14.2 1.07G| 1.8 16.9G 0.9 16.9G| | | 1 1
bitmapsearch16 R| 14.2 2.14G 7.1 2.14G| 1.8 16.9G 0.9 16.9G| | | 1 1
bitmapsearch32 R| 7.1 4.29G 3.5 4.35G| 1.8 16.9G 0.9 16.9G| | | 1 1
bitmapsearch64 R| 3.5 8.71G 1.8 8.47G| | | | 0 0
bitmapsearch8r R| 38.7 788M 19.1 798M| 3.0 10.1G 1.5 10.1G| | | 1 1
bitmapsearch16r R| 20.4 1.49G 10.0 1.52G| | | | 0 0
bitmapsearch32r R| 6.1 5.00G 3.2 4.76G| | | | 0 0
bitmapsearch64r R| 3.7 8.24G 2.0 7.62G| | | | 0 0
ChkSum32 R| 4.7 6.49G | 4.7 6.49G | 1.1 27.7G | | 2
All tests passed.
|
Processor data (only one entry):
processor : 0
vendor_id : GenuineIntel
cpu family : 6
model : 37
model name : Intel(R) Core(TM) i5 CPU 650 @ 3.20GHz
stepping : 5
cpu MHz : 1197.000
cache size : 4096 KB
physical id : 0
siblings : 4
core id : 0
cpu cores : 2
apicid : 0
initial apicid : 0
fpu : yes
fpu_exception : yes
cpuid level : 11
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx
rdtscp lm constant_tsc arch_perfmon pebs bts rep_good xtopology nonstop_tsc aperfmperf pni pclmulqdq dtes64 monitor ds_cpl vmx smx est tm2 ssse3 cx16 xtpr pdcm sse4_1 sse4_2 popcnt aes lahf_lm ida arat dts tpr_shadow vnmi flexpriority ept vpid
bogomips : 6399.26
clflush size : 64
cache_alignment : 64
address sizes : 36 bits physical, 48 bits virtual
power management:
|
Notes:
Intel H55 (shared video buffer), dual bank DDR3 1333 RAM 6-8-6-24
|
Go to short tests list |
|