Benchmark #744

Download
BORE vs BMQ vs SCX schedulers in Rocket League / kernel 6.11.8

Rocket League first two minutes of match replay no additional workload Ryzen 5700X RTX 3060Ti EPP powersave / balance_performance

Submitted 1 week ago by yubysowhat

Specifications
Label OS GPU CPU RAM OS specific
BMQ CachyOS NVIDIA GeForce RTX 3060 Ti AMD Ryzen 7 5700X 8-Core Processor 17 GB 6.11.8-1-cachyos-bmq-lto powersave
BORE CachyOS NVIDIA GeForce RTX 3060 Ti AMD Ryzen 7 5700X 8-Core Processor 17 GB 6.11.8-2-cachyos powersave
Bpfland CachyOS NVIDIA GeForce RTX 3060 Ti AMD Ryzen 7 5700X 8-Core Processor 17 GB 6.11.8-2-cachyos powersave
Flash CachyOS NVIDIA GeForce RTX 3060 Ti AMD Ryzen 7 5700X 8-Core Processor 17 GB 6.11.8-2-cachyos powersave
LAVD CachyOS NVIDIA GeForce RTX 3060 Ti AMD Ryzen 7 5700X 8-Core Processor 17 GB 6.11.8-2-cachyos powersave
Rusty CachyOS NVIDIA GeForce RTX 3060 Ti AMD Ryzen 7 5700X 8-Core Processor 17 GB 6.11.8-2-cachyos powersave
# Top runs: * **Highest FPS**: `BORE` offers the highest average FPS, approximately 2% better than the rest. * **Smoothest FPS**: `Flash` has the smoothest FPS with the lowest standard deviation, approximately 2% more stable than the others. * **Best overall**: `BORE` achieves the best overall performance by balancing the highest FPS and a fairly stable FPS. # Issues: * The configuration between the first scheduler and the others is different. The first run uses a slightly different kernel version. * The maximum CPU and GPU loads slightly vary across the runs, with up to a 2% difference in GPU loads. # Summary In this benchmark, different Linux kernel schedulers were compared while running a replay scenario in Rocket League on `CachyOS`. The setup included the Ryzen 5700X CPU and NVIDIA RTX 3060Ti GPU. The `BORE` scheduler demonstrated the highest average FPS, showing a tangible improvement over its counterparts. However, `Flash` provided the most consistent frame rate performance. Overall, the `BORE` scheduler offered the best balance of high performance and stability in this scenario. It's worth noting that the first scheduler was tested on a slightly distinct kernel version, which may have influenced the small variances observed.