Top runs:

• Highest FPS: bore with an average FPS approximately 2% higher than the next highest run. This indicates a robust performance in terms of frames per second.
• Smoothest FPS: bpfland_l with the lowest standard deviation and variance, approximately 2% lower than other runs, ensuring the smoothest frame timing and consistent performance.
• Best overall: bore as it achieves the highest FPS with a standard deviation that is only slightly higher, providing a balanced performance with both high FPS and relatively consistent frame timing.

Issues:

• The lavd_p run exhibited an average FPS approximately 1% lower than bore and 0.5% lower than bpfland_l. While this is not a significant drop, its higher variance compared to bpfland_l indicates a slightly less stable performance.
• The bpfland_l run had a higher Swap used average, around 94% compared to the others. This indicates potential inefficiencies in memory usage could impact longer runs.
• The lavd_p and bore runs had a slightly higher variance in the GPU core clock, around 4600% higher than bpfland_l.

Summary

This benchmark compares three different Rust schedulers on CachyOS using the same hardware with the performance Linux scheduler. The bore run delivers the highest average FPS, making it the top choice for raw performance. However, the bpfland_l run stands out as the smoothest, with the lowest frame timing variance, offering the most consistent gameplay experience. Despite being close in performance, the lavd_p run shows some minor instability and slightly higher GPU core clock variance, making it less ideal compared to the other two. Overall, when considering both FPS and smoothness, the bore run emerges as the best overall performer.