HankiPerformance Lab
Measure while you play with Hanki Performance Lab: the Monitor, fair comparisons, the Bottleneck Analyzer, Stutter Diagnostics and Performance sessions.
Performance Lab measures instead of guessing. Use it to see what limits a game and to check whether a change actually helped.
Tools
| Tab | What it does |
|---|---|
| Monitor | Measures every second: CPU and busiest thread, effective clock, memory, disk, GPU load, video memory, temperature and clock, network downloads and the busiest programs. With administrator rights it adds frame rates for DirectX games. |
| Comparisons | 30-second samples to compare before and after a change. |
| Bottleneck Analyzer | Names the likely limiter (GPU, CPU, memory, disk or background downloads) with its evidence and confidence. |
| Stutter Diagnostics | Matches frame-time spikes to CPU, GPU, memory, disk, heat and background activity, as possibilities. |
| Benchmarks | Planned: repeatable measurements for fair comparisons. |
| Advanced Tuning | Planned and kept separate: vendor GPU auto-tuning that always asks for explicit confirmation. |
Test a change
- Measure a baseline in the Monitor, or with Launch and measure in Gaming → Games.
- Apply one reviewed change.
- Measure again under the same conditions, then keep the change or restore it.
Hanki is conservative about the verdict. Without frame-rate data a run is only Measured. Runs must be comparable (same source, at least 75% of the length), and a change must improve by more than 3% without lowering 1% lows before Hanki calls it Improved.
Performance sessions
Runs and tests are saved in History → Performance sessions (the newest 100) with their baseline, the changes tested and the result. Saving a measurement to a file is your choice; measurements include the names of the busiest programs.
Updated · Found something out of date? Report it on GitHub.