LMBench2 Scalabilty

LMBench2 is a benchmark system that was designed to be portable between various *nix type systems including Linux.

LMBench2's tests have been loosely divided into 3 categories:

Configuration

Assumptions

Benefits

Problems

Results

cd results && make summary percent LIST=ia64-linux-gnu-first-run/* | ../scripts/lmbsum.pl 

                 L M B E N C H  2 . 0   S U M M A R Y
                 ------------------------------------

Processor, Processes - times in microseconds - smaller is better
----------------------------------------------------------------
                                 null     null                       open    signal   signal    fork    execve  /bin/sh
kernel                           call      I/O     stat    fstat    close   install   handle  process  process  process
-----------------------------  -------  -------  -------  -------  -------  -------  -------  -------  -------  -------
2.6.8-rc2-SMP                    0.214  0.29439   21.221    0.398   23.925    0.380    3.038    444.0   1304.8   3903.8
  s.d. (5 runs)                  0.000  0.00152    0.009    0.003    0.024    0.000    0.009      0.0      5.1      4.7
2.6.8-rc2-uni                    0.200  0.29439   21.316    0.389   24.207    0.398    3.119    449.0   1235.4   3809.2
  s.d. (5 runs)                  0.000  0.00090    0.006    0.006    0.010    0.002    0.005      0.0      1.7      3.2

File select - times in microseconds - smaller is better
-------------------------------------------------------
                                select   select   select   select   select   select   select   select
kernel                           10 fd   100 fd   250 fd   500 fd   10 tcp  100 tcp  250 tcp  500 tcp
-----------------------------  -------  -------  -------  -------  -------  -------  -------  -------
2.6.8-rc2-SMP                    1.893    7.015   15.643   30.464    2.438  12.1727  27.9824  56.0361
  s.d.                           0.014    0.173    0.414    0.372    0.012  0.11253  0.43376  0.59416
2.6.8-rc2-uni                    1.927    6.695   14.596   28.247    2.445  11.7909  27.4358  53.6381
  s.d.                           0.007    0.011    0.302    0.008    0.019  0.00814  0.00753  0.01739

Context switching with 0K - times in microseconds - smaller is better
---------------------------------------------------------------------
                                2proc/0k   4proc/0k   8proc/0k  16proc/0k  32proc/0k  64proc/0k  96proc/0k
kernel                         ctx swtch  ctx swtch  ctx swtch  ctx swtch  ctx swtch  ctx swtch  ctx swtch
-----------------------------  ---------  ---------  ---------  ---------  ---------  ---------  ---------
2.6.8-rc2-SMP                     10.154      6.870      6.008      5.924      6.430      5.612      5.336
  s.d.                             0.150      1.922      1.465      1.517      0.919      0.849      0.534
2.6.8-rc2-uni                      1.170      1.346      1.622      1.768      2.136      3.174      3.570
  s.d.                             0.010      0.015      0.022      0.018      0.033      0.055      0.070

Context switching with 4K - times in microseconds - smaller is better
---------------------------------------------------------------------
                                2proc/4k   4proc/4k   8proc/4k  16proc/4k  32proc/4k  64proc/4k  96proc/4k
kernel                         ctx swtch  ctx swtch  ctx swtch  ctx swtch  ctx swtch  ctx swtch  ctx swtch
-----------------------------  ---------  ---------  ---------  ---------  ---------  ---------  ---------
2.6.8-rc2-SMP                      8.778      8.078      7.358      6.810      7.144      6.846      6.978
  s.d.                             3.791      2.313      2.379      2.263      2.294      1.142      0.953
2.6.8-rc2-uni                      1.540      1.812      2.104      2.352      2.838      4.282      5.610
  s.d.                             0.017      0.055      0.040      0.013      0.050      0.134      0.197

Context switching with 8K - times in microseconds - smaller is better
---------------------------------------------------------------------
                                2proc/8k   4proc/8k   8proc/8k  16proc/8k  32proc/8k  64proc/8k  96proc/8k
kernel                         ctx swtch  ctx swtch  ctx swtch  ctx swtch  ctx swtch  ctx swtch  ctx swtch
-----------------------------  ---------  ---------  ---------  ---------  ---------  ---------  ---------
2.6.8-rc2-SMP                     10.858      8.406      9.344      9.100      8.690      8.576      8.288
  s.d.                             0.144      2.413      1.853      2.163      1.665      1.874      1.200
2.6.8-rc2-uni                      1.808      2.120      2.428      2.790      3.324      4.714      6.224
  s.d.                             0.004      0.043      0.025      0.010      0.011      0.071      0.265

Context switching with 16K - times in microseconds - smaller is better
----------------------------------------------------------------------
                               2proc/16k  4proc/16k  8proc/16k  16prc/16k  32prc/16k  64prc/16k  96prc/16k
kernel                         ctx swtch  ctx swtch  ctx swtch  ctx swtch  ctx swtch  ctx swtch  ctx swtch
-----------------------------  ---------  ---------  ---------  ---------  ---------  ---------  ---------
2.6.8-rc2-SMP                      9.622      7.948      6.890      7.220      7.670      8.406      8.962
  s.d.                             3.810      1.998      1.723      2.071      1.906      0.994      1.290
2.6.8-rc2-uni                      2.348      2.678      3.042      3.948      4.374      6.014      7.610
  s.d.                             0.013      0.033      0.019      0.011      0.070      0.150      0.170

Context switching with 32K - times in microseconds - smaller is better
----------------------------------------------------------------------
                               2proc/32k  4proc/32k  8proc/32k  16prc/32k  32prc/32k  64prc/32k  96prc/32k
kernel                         ctx swtch  ctx swtch  ctx swtch  ctx swtch  ctx swtch  ctx swtch  ctx swtch
-----------------------------  ---------  ---------  ---------  ---------  ---------  ---------  ---------
2.6.8-rc2-SMP                      9.904      9.178     10.442     10.568     11.120     11.142     11.888
  s.d.                             3.781      2.003      1.762      0.910      1.147      0.989      1.054
2.6.8-rc2-uni                      3.386      3.888      5.104      6.014      6.376      8.730     13.462
  s.d.                             0.005      0.036      0.024      0.011      0.111      0.149      0.412

Context switching with 64K - times in microseconds - smaller is better
----------------------------------------------------------------------
                               2proc/64k  4proc/64k  8proc/64k  16prc/64k  32prc/64k  64prc/64k  96prc/64k
kernel                         ctx swtch  ctx swtch  ctx swtch  ctx swtch  ctx swtch  ctx swtch  ctx swtch
-----------------------------  ---------  ---------  ---------  ---------  ---------  ---------  ---------
2.6.8-rc2-SMP                     14.500     13.462     14.590     16.036     15.778     16.386     20.234
  s.d.                             0.007      2.166      0.962      0.971      0.414      0.543      1.780
2.6.8-rc2-uni                      5.618      8.060      9.340      9.710     10.530     21.736     56.858
  s.d.                             0.073      0.021      0.010      0.019      0.213      2.238      4.236

File create/delete and VM system latencies in microseconds - smaller is better
----------------------------------------------------------------------------
                                 0K       0K       1K       1K       4K       4K      10K      10K     Mmap     Prot    Page
kernel                         Create   Delete   Create   Delete   Create   Delete   Create   Delete   Latency  Fault   Fault------------------------------ -------  -------  -------  -------  -------  -------  -------  -------  -------  ------  ------
2.6.8-rc2-SMP                    52.46    34.50    74.15    46.71    74.26    46.45    90.70    50.67   3428.2   0.582   34.40
  s.d.                            0.02     0.07     0.67     0.14     0.70     0.35     0.94     0.68     15.8   0.016   32.21
2.6.8-rc2-uni                    54.18    36.46    75.53    48.25    75.78    48.35    92.40    52.61   3174.0   0.667   69.20
  s.d.                            0.30     0.02     0.07     0.27     0.24     0.26     0.63     0.48     35.7   0.020   29.90

*Local* Communication latencies in microseconds - smaller is better
-------------------------------------------------------------------
kernel                           Pipe   AF/Unix     UDP   RPC/UDP     TCP   RPC/TCP  TCPconn
-----------------------------  -------  -------  -------  -------  -------  -------  -------
2.6.8-rc2-SMP                  101.492   45.287  59.9310  44.7041  197.394  213.471   83.373
  s.d.                          11.710    1.666  22.5623  0.46736  1.12974  2.36252    0.632
2.6.8-rc2-uni                    4.463   13.374  16.8566  25.9673  180.202  192.352  116.399
  s.d.                           0.021    0.016  0.06387  0.25551  1.80836  1.87359    0.263

*Local* Communication bandwidths in MB/s - bigger is better
-----------------------------------------------------------
                                                             File     Mmap    Bcopy    Bcopy   Memory   Memory
kernel                           Pipe   AF/Unix    TCP     reread   reread   (libc)   (hand)     read    write
-----------------------------  -------  -------  -------  -------  -------  -------  -------  -------  -------
2.6.8-rc2-SMP                  3002.92   164.04   173.59  1629.78   780.56   730.94   486.42   780.63   667.59
  s.d.                           16.53     5.60     1.02     1.88     0.38    12.30     1.81     0.42     0.46
2.6.8-rc2-uni                  3071.89   184.73   178.29  1629.80   780.46   735.13   486.99   780.49   666.14
  s.d.                           10.05     0.13     0.09     2.22     0.06    10.61     1.65     0.09     0.06

*Local* More Communication bandwidths in MB/s - bigger is better
----------------------------------------------------------------
                                  File     Mmap  Aligned  Partial  Partial  Partial  Partial
OS                                open     open    Bcopy    Bcopy     Mmap     Mmap     Mmap    Bzero
                                 close    close   (libc)   (hand)     read    write   rd/wrt     copy     HTTP
-----------------------------  -------  -------  -------  -------  -------  -------  -------  -------  -------
2.6.8-rc2-SMP                  1629.63   738.06   733.26   735.67   951.27  1585.75   645.77  2253.45    6.134
  s.d.                            2.83     0.74    14.69    13.60     0.45     2.91     0.40     6.79    0.080
2.6.8-rc2-uni                  1629.36   740.57   735.50   735.09   951.22  1579.64   642.05  2245.85    6.034
  s.d.                            0.92     0.50    11.69    13.37     0.13     0.50     0.18     3.68    0.232

Memory latencies in nanoseconds - smaller is better
---------------------------------------------------
kernel                          Mhz     L1 $     L2 $    Main mem
-----------------------------  -----  -------  -------  ---------
2.6.8-rc2-SMP                   1501    1.335    4.010      97.53
  s.d.                             0    0.080    0.080       0.08
2.6.8-rc2-uni                   1501    1.335    4.006      97.71
  s.d.                             0    0.232    0.232       0.23

IA64wiki: LMBench2 Scale (last edited 2009-12-10 03:13:46 by localhost)

Gelato@UNSW is sponsored by
the University of New South Wales National ICT Australia The Gelato Federation Hewlett-Packard Company Australian Research Council
Please contact us with any questions or comments.