Some Unix benchmarks for those who are interesed
Artem Kuchin
matrix at itlegion.ru
Tue Mar 6 08:48:06 UTC 2007
I used unixbenchmark for measure overall perfomance of three machines of
different generations and with different OS versions. See for your self and
compare results with machine cost.
Hope this will be usefull for someone.
BENCHMARKS:
OMNI3 (current cost 2400$):
Intel entry server mainboard, intel 1U case,
20 2HDD MIRROR, 2GB ECC 533Mhz, Pentium D 3.4Ghz
Built-in LSI MegaRAID RAID controller on the mainbopard.
FreeBSD 6.2-STABLE
/bin/sh: ELF 64-bit LSB executable, AMD x86-64, version 1 (FreeBSD), dynamically linked (uses shared libs), stripped
/dev/ar0s1e 289385582 2514874 263719862 1% /usr
Dhrystone 2 using register variables 10588607.7 lps (10.0 secs, 10 samples)
Double-Precision Whetstone 1293.5 MWIPS (10.6 secs, 10 samples)
System Call Overhead 494540.4 lps (10.0 secs, 10 samples)
Pipe Throughput 597906.5 lps (10.0 secs, 10 samples)
Pipe-based Context Switching 34342.0 lps (12.3 secs, 10 samples)
Process Creation 3009.9 lps (38.8 secs, 3 samples)
Execl Throughput no measured results
File Read 1024 bufsize 2000 maxblocks 464078.0 KBps (30.0 secs, 3 samples)
File Write 1024 bufsize 2000 maxblocks 110822.0 KBps (30.0 secs, 3 samples)
File Copy 1024 bufsize 2000 maxblocks 110268.0 KBps (30.0 secs, 3 samples)
File Read 256 bufsize 500 maxblocks 124661.0 KBps (30.0 secs, 3 samples)
File Write 256 bufsize 500 maxblocks 86272.0 KBps (
File Copy 256 bufsize 500 maxblocks 50327.0 KBps (
File Read 4096 bufsize 8000 maxblocks 1394205.0 KBps
File Write 4096 bufsize 8000 maxblocks 67200.0 KBps (
File Copy 4096 bufsize 8000 maxblocks 64455.0 KBps (
Shell Scripts (1 concurrent) 3212.6 lpm (
Shell Scripts (8 concurrent) 579.2 lpm (
Shell Scripts (16 concurrent) 293.2 lpm (
Arithmetic Test (type = short) 892425.3 lps (
Arithmetic Test (type = int) 954128.2 lps (
Arithmetic Test (type = long) 298186.4 lps (
Arithmetic Test (type = float) 1035063.3 lps
Arithmetic Test (type = double) 516773.0 lps (
Arithoh 17328248.7 lps
C Compiler Throughput 1815.1 lpm (
Dc: sqrt(2) to 99 decimal places 73558.8 lpm (
Recursion Test--Tower of Hanoi 141070.4 lps (
INDEX VALUES
TEST BASELINE
Dhrystone 2 using register variables 116700.0 1058
Double-Precision Whetstone 55.0
Execl Throughput 43.0
File Copy 1024 bufsize 2000 maxblocks 3960.0 11
File Copy 256 bufsize 500 maxblocks 1655.0 5
File Copy 4096 bufsize 8000 maxblocks 5800.0 6
Pipe Throughput 12440.0 59
Pipe-based Context Switching 4000.0 3
Process Creation 126.0
Shell Scripts (8 concurrent) 6.0
System Call Overhead 15000.0 49
OMNI2 (current cost approx 4000$):
Supermicro mainboard, supermicro case.
360 3HDD RAID-5, 4GB ECC 400Mhz, DUal Xeon 2.67Ghz
3WARE 8xxxx with 4 PORTS
FreeBSD 5.4-STABLE
/bin/sh: ELF 32-bit LSB executable, Intel 80386, version 1 (FreeBSD), for FreeBSD 5.4-CURRENT (rev 2), dynamically linked (
uses shared libs), stripped
/dev/twed0s1e 10154158 5684520 3657306 61% /usr
Dhrystone 2 using register variables 3979893.5 lps (10.0 secs, 10 samples)
Double-Precision Whetstone 837.4 MWIPS (10.3 secs, 10 samples)
System Call Overhead 218348.2 lps (10.0 secs, 10 samples)
Pipe Throughput 236742.1 lps (10.1 secs, 10 samples)
Pipe-based Context Switching 56205.9 lps (10.0 secs, 10 samples)
Process Creation 2662.7 lps (30.0 secs, 3 samples)
Execl Throughput 719.6 lps (29.6 secs, 3 samples)
File Read 1024 bufsize 2000 maxblocks 250541.0 KBps (30.0 secs, 3 samples)
File Write 1024 bufsize 2000 maxblocks 11986.0 KBps (30.0 secs, 3 samples)
File Copy 1024 bufsize 2000 maxblocks 12850.0 KBps (30.0 secs, 3 samples)
File Read 256 bufsize 500 maxblocks 71142.0 KBps (30.0 secs, 3 samples)
File Write 256 bufsize 500 maxblocks 10535.0 KBps (30.0 secs, 3 samples)
File Copy 256 bufsize 500 maxblocks 10247.0 KBps (
File Read 4096 bufsize 8000 maxblocks 617670.0 KBps (
File Write 4096 bufsize 8000 maxblocks 26746.0 KBps (
File Copy 4096 bufsize 8000 maxblocks 21936.0 KBps (
Shell Scripts (1 concurrent) 1285.4 lpm (
Shell Scripts (8 concurrent) 163.4 lpm (
Shell Scripts (16 concurrent) 133.2 lpm (
Arithmetic Test (type = short) 630290.8 lps (
Arithmetic Test (type = int) 601545.1 lps (
Arithmetic Test (type = long) 627739.3 lps (
Arithmetic Test (type = float) 670343.5 lps (
Arithmetic Test (type = double) 664330.5 lps (
Arithoh 13388181.7 lps
C Compiler Throughput 1159.6 lpm (
Dc: sqrt(2) to 99 decimal places 74973.1 lpm (
Recursion Test--Tower of Hanoi 46712.1 lps (
INDEX VALUES
TEST BASELINE
Dhrystone 2 using register variables 116700.0 397
Double-Precision Whetstone 55.0
Execl Throughput 43.0
File Copy 1024 bufsize 2000 maxblocks 3960.0 1
File Copy 256 bufsize 500 maxblocks 1655.0 1
File Copy 4096 bufsize 8000 maxblocks 5800.0 2
Pipe Throughput 12440.0 23
Pipe-based Context Switching 4000.0 5
Process Creation 126.0
Shell Scripts (8 concurrent) 6.0
System Call Overhead 15000.0 21
OSIRIS (current cost: hm.. probably about 700$ beacuse of 3ware controller and HHDs)
Self assembled, gigabyte mainboard, cheap medium tower case + 100$ PSU.
240 3 HDD RAID-5, 756MB ECC SRAM 133Mhz, Dual Pentium 3 550Mhz
3ware 7xxxx series for 4 ports.
FreeBSD 6.1-STABLE
/bin/sh: ELF 32-bit LSB executable, Intel 80386, version 1 (FreeBSD), dynamically linked (uses shared libs), stripped
/dev/twed0s1e 20308398 10469660 8214068 56% /usr
Dhrystone 2 using register variables 1230939.8 lps (10.0 secs, 10 samples)
Double-Precision Whetstone 332.7 MWIPS (10.1 secs, 10 samples)
System Call Overhead 149582.4 lps (10.0 secs, 10 samples)
Pipe Throughput 158045.2 lps (10.0 secs, 10 samples)
Pipe-based Context Switching 19863.0 lps (10.0 secs, 10 samples)
Process Creation 938.3 lps (30.0 secs, 3 samples)
Execl Throughput 321.9 lps (29.7 secs, 3 samples)
File Read 1024 bufsize 2000 maxblocks 106492.0 KBps (30.0 secs, 3 samples)
File Write 1024 bufsize 2000 maxblocks 9555.0 KBps (30.0 secs, 3 samples)
File Copy 1024 bufsize 2000 maxblocks 10572.0 KBps (30.0 secs, 3 samples)
File Read 256 bufsize 500 maxblocks 35077.0 KBps (30.0 secs, 3 samples)
File Write 256 bufsize 500 maxblocks 7663.0 KBps (
File Copy 256 bufsize 500 maxblocks 8480.0 KBps (
File Read 4096 bufsize 8000 maxblocks 236555.0 KBps (
File Write 4096 bufsize 8000 maxblocks 12886.0 KBps (
File Copy 4096 bufsize 8000 maxblocks 10266.0 KBps (
Shell Scripts (1 concurrent) 770.2 lpm (
Shell Scripts (8 concurrent) 128.2 lpm (
Shell Scripts (16 concurrent) 64.0 lpm (
Arithmetic Test (type = short) 137979.8 lps (
Arithmetic Test (type = int) 138354.4 lps (
Arithmetic Test (type = long) 138406.5 lps (
Arithmetic Test (type = float) 172898.2 lps (
Arithmetic Test (type = double) 172916.2 lps (
Arithoh 2551887.3 lps
C Compiler Throughput 332.0 lpm (
Dc: sqrt(2) to 99 decimal places 23695.0 lpm (
Recursion Test--Tower of Hanoi 20153.3 lps (
INDEX VALUES
TEST BASELINE
Dhrystone 2 using register variables 116700.0 123
Double-Precision Whetstone 55.0
Execl Throughput 43.0
File Copy 1024 bufsize 2000 maxblocks 3960.0 1
File Copy 256 bufsize 500 maxblocks 1655.0
File Copy 4096 bufsize 8000 maxblocks 5800.0 1
Pipe Throughput 12440.0 15
Pipe-based Context Switching 4000.0 1
Process Creation 126.0
Shell Scripts (8 concurrent) 6.0
System Call Overhead 15000.0 14
--
Regards,
Artem
More information about the freebsd-stable
mailing list