Benchmark #1232

Download
Bpfland in Rocket League | kernel 6.15rc1

Rocket League | first two minutes of match replay | no additional workload | Ryzen 5700X | RTX 3060Ti | power profile balanced

Submitted 1 week ago by yubysowhat

Specifications
Label OS GPU CPU RAM OS specific
Bpfland-next CachyOS NVIDIA GeForce RTX 3060 Ti AMD Ryzen 7 5700X 8-Core Processor 34 GB 6.15.0-rc1-1-cachyos-rc powersave
Bpfland-next -m 0xff CachyOS NVIDIA GeForce RTX 3060 Ti AMD Ryzen 7 5700X 8-Core Processor 34 GB 6.15.0-rc1-1-cachyos-rc powersave
Bpfland-next -n CachyOS NVIDIA GeForce RTX 3060 Ti AMD Ryzen 7 5700X 8-Core Processor 34 GB 6.15.0-rc1-1-cachyos-rc powersave
Bpfland-next -n -k CachyOS NVIDIA GeForce RTX 3060 Ti AMD Ryzen 7 5700X 8-Core Processor 34 GB 6.15.0-rc1-1-cachyos-rc powersave
Bpfland-next -n -p CachyOS NVIDIA GeForce RTX 3060 Ti AMD Ryzen 7 5700X 8-Core Processor 34 GB 6.15.0-rc1-1-cachyos-rc powersave
# Top runs: * **Highest FPS**: `Bpfland-next -m 0xff` with a slightly higher average FPS than the rest. * **Smoothest FPS**: `Bpfland-next -m 0xff` with the lowest standard deviation and variance in FPS. * **Best overall**: `Bpfland-next -m 0xff` providing a good balance of high FPS and smoothness. # Issues: * The runs labeled with `-n`, `-n -k`, and `-n -p` have significantly lower FPS, around 12% less than the highest run. This indicates a drop in performance with that configuration. * Standard deviation in the FPS for `-n`, `-n -k`, and `-n -p` is substantially higher, suggesting less stability in these runs. # Summary In this benchmark, the performance of Rocket League was evaluated under different configurations using the `Bpfland` scheduler on `CachyOS` with a `powersave` scheduling policy. The configuration tagged as `-m 0xff` delivered the best performance, achieving the highest average FPS and the most stable frame rate. However, configurations with `-n`, `-n -k`, and `-n -p` experienced a notable decrease in FPS and stability, indicating that adjustments with this setup can substantially impact performance.