"micbench" is a set of programs for measuring basic performance of your machines.

• micbench-io : I/O performance
• micbench-mem : Memory access performance

• Ruby interpreter (1.8.7 or later)
• numactl
• libnuma

Currently it also requires Nehalem or newer x86_64 architectures.

$ ./configure
$ make

$ micbench mem -m 1 -t 10 -R -a 0:c0 -s 16K -v
access_pattern  random
multiplicity    1
local   true
page_size       4096
size    16384
use_hugepages   false
total_ops       5602803712
total_clk       22561968928
exec_time       10.000461
ops_per_sec     5.602545e+08
clk_per_op      4.026907e+00
total_exec_time 11.518878

• -m 1 ... with 1 thread
• -t 10 ... the benchmark runs for 10 seconds
• -R ... random access mode
• -a 0:c0 ... a thread 0 is executed only on cpu0
• -s 16K ... the size of memory region accessed is 16KB
• -v ... verbose mode

The result clk_per_op shows how many clocks is needed to access the L1 cache of your processor.

In random access mode of micbench-mem, each memory access operation has data dependency on preceding operation, so all the operations are totally serialized in a pipeline of a processor. It means that dividing execution time by # of memory access operations is almost equal to the memory access latency of the processor. In this case, only 16KB of memory is accessed and all memory access hits the L1 cache. Therefore the results shows the access latency of the L1 cache.

Latest source code is on GitHub: http://github.com/hayamiz/micbench/

You can build micbench from scratch by executing commands below. In addition to prerequisites above, you need autotools to build.

$ ./autogen.sh
$ ./configure
$ make

If you want to run tests, you also need GLib and Cutter.

• GLib : http://www.gtk.org/
• Cutter : http://cutter.sf.net/