zephyr benchmark sample

Browse source code on GitHub

Overview

This sample demonstrates the usage of Zephyr Benchmarking subsystem to measure performance of code blocks reliably.

Sample Output

This is the sample output when running the sample application on qemu_x86:

math_benchmarks ############################################
[00:00:00.160,000] <dbg> test_benchmarks: suit_setup: Setting up math benchmarks suite
benchmark_addition =========================================
   Sample size:100, total cycles: 28800
   Mean(μ̂): 288.000
   Standard deviation(s): 0.000
   Standard Error(SE): 0.000
   Min: 288 (run #1)
   Max: 288 (run #1)
benchmark_pure_asm_add =====================================
   Sample size:100, total cycles: 3200
   Mean(μ̂): 32.000
   Standard deviation(s): 0.000
   Standard Error(SE): 0.000
   Min: 32 (run #1)
   Max: 32 (run #1)
benchmark_subtraction ======================================
   Sample size:100, total cycles: 32000
   Mean(μ̂): 320.000
   Standard deviation(s): 0.000
   Standard Error(SE): 0.000
   Min: 320 (run #1)
   Max: 320 (run #1)
void_test ==================================================
   Sample size:100, total cycles: 0
   Mean(μ̂): 0.000
   Standard deviation(s): 0.000
   Standard Error(SE): 0.000
   Min: 0 (run #1)
   Max: 0 (run #1)
[00:00:00.200,000] <dbg> test_benchmarks: suit_teardown: Tearing down math benchmarks suite