# Top runs: * **Highest FPS**: `BPFLAND-1ms` – This run delivered the highest average frame rate with a mean FPS approximately 1% higher than the next closest run. * **Smoothest FPS**: `BPFLAND-20ms` – This run exhibited the lowest standard deviation and variance in FPS, indicating the most stable and consistent performance. * **Best overall**: `BPFLAND-1ms` – This run combines the highest average FPS with relatively low standard deviation and variance, providing both high and consistent performance. # Issues: * The run with a `20ms` slice size has significantly lower FPS, with its mean FPS around 15% lower than the best run. * The run with a `10ms` slice size also shows reduced performance with an approximately 9% lower mean FPS compared to the top performance run. * The runs with `15ms` slice size and higher exhibit a noticeable decline in performance metrics compared to the runs with smaller slice sizes. # Summary This benchmark compares the performance of different `BPFLAND` slice sizes using `CarX Drift Racing Online` under a `YouTube 1080p30 foreground load`. The results show that smaller slice sizes like `0.5ms` and `1ms` provide the best FPS performance, while maintaining relatively low frame time variability. However, as the slice size increases (particularly starting from `10ms`), there is a consistent drop in FPS, indicating lower performance. The smallest slice size tested, `20ms`, resulted in the most stable FPS but also the lowest mean FPS, highlighting a trade-off between stability and raw performance. All benchmarks were conducted under consistent hardware and software conditions, ensuring the differences are attributable to the slice size variations.