DOKK / manpages / debian 11 / likwid / likwid-bench.1.en
LIKWID-BENCH(1) General Commands Manual LIKWID-BENCH(1)

likwid-bench - low-level benchmark suite and microbenchmarking framework

likwid-bench [-hap] [-t <testname>] [-s <min_time>] [-w <workgroup_expression>] [-W <workgroup_expression_short>] [-l <testname>] [-d <delimiter>] [-i <iterations>] [-f <filepath>]

likwid-bench is a benchmark suite for low-level (assembly) benchmarks to measure bandwidths and instruction throughput for specific instruction code on x86 systems. The currently included benchmark codes include common data access patterns like load and store but also calculations like vector triad and sum. likwid-bench includes architecture specific benchmarks for x86, x86_64 and x86 for Intel Xeon Phi coprocessors. With LIKWID 5 also ARM and POWER benchmarks are supported. The performance values can either be calculated by likwid-bench or measured using performance counters by using likwid-perfctr as a wrapper to likwid-bench. This requires to build likwid-bench with instrumentation enabled in config.mk. Benchmarks can be dynamically added when a proper ptt file is present at $HOME/.likwid/bench/<arch>/<testname>.ptt . The files are compiled to a .S file and compiled using either gcc, icc or pgcc (searched in $PATH). The default folder is /tmp/<PID>. Possible values for <arch> are 'x86', 'x86-64', 'phi', armv7', 'armv8' and 'power'.

prints a help message to standard output, then exits.
list available benchmark codes for the current system.
list available thread domains.
Run the benchmark for at least <min_time> seconds. The amount of iterations is determined using this value. Default: 1 second.
Name of the benchmark code to run (mandatory).
Specify the affinity domain, thread count and data set size for the current benchmarking run (-w or -W mandatory). First thread in thread domain initializes the stream.
Specify the affinity domain, thread count and data set size for the current benchmarking run (-w or -W mandatory). Each thread in the workgroup initializes its own chunk of the stream.
list properties of a benchmark code.
Set the number of iterations per thread (optional)
Filepath for the dynamic generation of benchmarks. Default /tmp/. <PID> is always attached

<thread_domain>:<size> [:<num_threads>[:<chunk_size>:<stride>]] [-<streamId>:<domain_id>] with size in kB, MB or GB. The <thread_domain> defines where the threads are placed. <size> is the total data set size for the benchmark, the allocated vectors in memory sum up to this size. <num_threads> specifies how many threads are used in the <thread_domain>. Threads are always placed using a compact policy in likwid-bench. This means that per default all SMT threads are used. Optionally similar a the expression based syntax in likwid-pin a <chunk_size> and <stride> can be provided. Optionally for every stream (array, vector) the placement can be controlled. Per default all arrays are placed in the same <thread_domain> the threads are running in. To place the data in a different domain for every stream of a benchmark case (the total number of streams can be acquired by the -l option) the domain to place the data in can be specified. Multiple streams are comma separated. Either the placement is provided or all streams have to be explicitly placed. Please refer to the Wiki pages on https://github.com/RRZE-HPC/likwid/wiki/Likwid-Bench for further details and examples on usage. With -W each thread initializes its own chunk of the streams but pleacement of the streams is deactivated.

1.
Run the copy benchmark on socket 0 ( S0 ) with a total data set size of 100kB.

Since no <num_threads> is given in the workload expression, each hardware thread of socket 0 gets one application thread. The workload is split up between all threads and the number of iterations is determined automatically.

2.
Run the triad benchmark code with explicitly 100 iterations per thread with 2 threads on the socket 0 ( S0 ) and a data size of 1GB.

Assuming socket 0 ( S0 ) has 2 physical hardware threads with SMT enabled, hence in total 4 hardware threads, one thread is assigned to each physical hardware thread of socket 0.

3.
Run the update benchmark on socket 0 ( S0 ) with a workload of 100kB and on socket 1 ( S1 ) with the same workload.

The results of both workgroups are combinded for the output. Hence the workload in each workgroup expression should have the same size.

4.
Run the copy benchmark but measure the memory traffic with likwid-perfctr. The option INSTRUMENT_BENCH in config.mk needs to be true at compile time to use that feature.

likwid-perfctr will configure and start the performance counters on socket 0 ( S0 ) with 4 threads prior to the execution of likwid-bench. The performance counters are read right before and after running the benchmarking code to minimize the interferences of the measurement.

5.
Run the copy benchmark and place the data on another socket

Stream id 0 and 1 are placed in thread domains S1, which is socket 1. This can be verified as the initialization threads output where they are running.

Since LIKWID 5.0, it is possible to have different numbers of threads in workgroups. Also different sizes are allowed. Both features seem promising, but they show a range of problems. If you have a NUMA system and run with multiple threads on NUMA node 0 but with less on NUMA node 1, the threads on NUMA node 1 cause less preassure on the memory interface and consequently achieve higher throughput. They will finish early compared to the threads on NUMA node 0. The runtime used for caluclating the bandwidth and MFlops/s values use the maximal runtime of all threads, hence one of NUMA node 0. Similar problems exist with different sizes. One workgroup might run in cache while the other waits for data from the memory interface.

Written by Thomas Gruber <thomas.roehl@googlemail.com>.

Report Bugs on <https://github.com/RRZE-HPC/likwid/issues>.

likwid-perfctr(1), likwid-pin(1), likwid-topology(1), likwid-setFrequencies(1)

16.11.2020 likwid-5