Tools
Benchmark and how to measure properly
Run same-duration before and after measurements and read the significance verdicts.
Changes only matter if you can measure them. The Benchmark under Gaming › Benchmark records how your PC behaves for a fixed time, so you can compare before and after.
What it measures
At a fixed interval for a fixed duration: CPU use, available memory, disk activity, GPU use, process count and network latency. Each run also records your configuration (Windows, CPU, GPU, power plan) and which Nexus changes were active.
A good before and after
- Baseline: close what you normally close, then run the benchmark. Use the same duration every time.
- Apply your changes. Restart if the result asks you to.
- After: recreate the same conditions and run again with the same duration.
- If you test a game, use its built-in benchmark and enter the FPS for both runs.
- Compare the two runs.
Reading the verdicts
Each metric is marked Improved, Worse or No significant change. A change only counts when it is clearly bigger than the run-to-run noise (a Welch's t-test) and larger than a minimum real-world difference. No significant change is a perfectly normal result, and an honest one: it means the change didn't measurably matter on this PC.
If a tweak with Limited evidence shows no improvement, undo it in Restore Center.