# Top runs: * **Highest FPS**: `OR - bprland Ini Tweak Ultra Perf` with an average FPS of 98.24%, which is around 25% higher than the other runs. * **Smoothest FPS**: `OR - bpfland Stock Hardware Lumen Ultra Perf` with a standard deviation of 8.997% and variance of 80.954%, which is significantly lower than the other runs. * **Best overall**: `OR - bpfland Stock Hardware Lumen Ultra Perf` as it has a good balance of high FPS (around 99% of the highest run) and very smooth frame times. # Issues: * The `OR - bprland Ini Tweak Ultra Perf` run has significantly higher FPS compared to the other runs, but this comes at the cost of higher CPU and GPU load, as well as higher temperatures. This suggests that this configuration may not be sustainable in the long run and could potentially lead to performance degradation or system instability. # Summary The benchmark compares the performance of the Oblivion Remaster game on a system with an AMD Ryzen 9 9950X3D CPU and an NVIDIA GeForce RTX 5080 GPU, running on the Steam Runtime 3 (sniper) environment with Linux kernel 6.14.4-pikaos. The key focus of the comparison is the use of the `bpfland` Linux scheduler, which is a custom scheduler implemented using the `sched_ext` feature. The results show that the `bpfland` scheduler can provide a significant improvement in terms of FPS smoothness, as seen in the `OR - bpfland Stock Hardware Lumen Ultra Perf` run. This run has the lowest standard deviation and variance in FPS, indicating a more stable and consistent frame delivery. However, the `OR - bprland Ini Tweak Ultra Perf` run, which does not use the `bpfland` scheduler, manages to achieve the highest average FPS, around 25% higher than the other runs. The main tradeoff seems to be between FPS stability and raw performance. The `bpfland` scheduler prioritizes smoothness, while the other configurations focus more on maximizing the peak FPS. Depending on the user's preferences and the specific requirements of the game, either of these approaches could be considered the "best overall" option.