Test method
Standard procedure (all sims)
- Use a repeatable scenario: a saved replay of a full-grid race is best, because traffic and cameras are identical every run. Otherwise an AI race start from a fixed grid slot.
- Same car, track, time of day, weather and car count for every run. Note them in the form.
- Triples: write down the combined render resolution (e.g. 7680x1440), whether you use the sim's own triple rendering or NVIDIA Surround / AMD Eyefinity, and in iRacing whether SMP is on.
- Turn off overlays you do not need. Note any frame cap: capped results are shown separately.
- Capture at least 60 seconds with CapFrameX or PresentMon (both MIT-licensed). VR: XRFrameTools.
- Run it three times and submit the middle run. Upload the file and a screenshot of your graphics settings.
Method draws on pixelsonly.racing, OverTake.gg and Tom's Hardware.
What we calculate
Average FPS = frames ÷ total capture time. 1% low = average of the slowest 1% of frames, as FPS. P1 = 99th-percentile frame time, as FPS. The server recalculates all three from your file.