Media experiments

Benchmark comparison

Run benchmarks

Run each selected experiment in this browser, one at a time. Keep this page visible and close other experiment players while measuring.

Experiments

Source and network controls affect every viewer of this deployment. Repeated comparisons rotate the experiment order.

Loading services…

Compare selected runs

Browser CPU and total browser memory require a CLI result import. CPU 100% means one core. Backend process measurements exclude source encoding and container totals. Missing measurements are shown with their availability reason.

Performance measurements for selected runs. Cumulative counters use the maximum.

Saved runs

Select up to nine runs to compare. Imports stay in this page; collector exports persist on the server.

Most recent 40 completed collector runs and imported results
CompareExperiment / startedScenario / sourceResultVersion / exports