A MaximusFPS guide
What are 1% lows and 0.1% lows, and how can I improve them?
1% lows and 0.1% lows describe the slow end of frame delivery rather than the average. They help reveal brief slowdowns that a high average FPS can hide. Calculation methods vary between tools, so compare results from the same tool and test method, and inspect the frametime graph instead of relying on one summary number.
What to check first
- Use the same tool, capture length, game scene and settings for every comparison.
- Warm up the game and repeat runs; a loading screen can distort the result.
- Record clocks, temperatures and memory use when the slowest frames occur.
Likely causes and how to distinguish them
Shader compilation, asset loading, CPU scheduling and memory pressure can add occasional long frames. A 0.1% result is especially sensitive to a small number of outliers and short captures.
Average FPS describes overall throughput. Low-percentile results describe the slow tail, while a frametime graph shows the timing pattern. None of these measures directly proves mouse-to-display latency or network quality.
Diagnostic steps, in order
- Choose a representative route with similar movement and effects. Exclude menus and loading screens consistently, not only from the better result.
- Run multiple captures and compare the spread. If results vary widely without changes, investigate the test conditions before claiming a gain.
- Test background activity and game settings separately. Reduce settings that stress the limiting component rather than simply lowering every option.
- Check memory configuration and cooling only when measurements point there. Return unstable tuning to stock and repeat the test before judging performance.
Limitations and when help may be appropriate
Not every long frame is user-fixable; engine behaviour and first-run work may remain. Help is useful when the graph reveals a pattern you cannot interpret. Do not promise an improvement in lows from a memory profile or a service without measured evidence on that system.
Follow-up questions
Is a 1% low the minimum FPS?
No. It is a summary of a slow portion of the capture, depending on the tool’s calculation. The minimum may represent a single isolated frame and can be much less representative.
Can I compare results from different tools?
Be cautious. Tool definitions and capture conditions may differ. Use a consistent tool and repeatable workload before comparing changes.
Official sources and further reading
These documents support the technical explanations; game and driver options can change. Source language may differ from this guide.