Benchmark #1886

Download
Nine Sols: bpfland-next -S

GPU Limited static scene + CPU Encoded 1080p60 Discord Stream | 30 second frame recording with 0ms interval | Performance Governor/EPP | Proton GE 10-24 Wayland Native | SCX Next: e9e2b21 | kernel-cachyos-lto-6.17.1-cachyos1.lto.fc44.x86_64 with Staging SCX Patches.

Submitted 5 days ago by summedkibbles36

Specifications
Label OS GPU CPU RAM OS specific
scx_bpfland | 1 Steam Runtime 3 (sniper) AMD Radeon RX 7900 GRE (RADV NAVI31) AMD Ryzen 7 7800X3D 8-Core Processor 33 GB 6.17.1-cachyos1.lto.fc44.x86_64 performance
scx_bpfland -S | 1 Steam Runtime 3 (sniper) AMD Radeon RX 7900 GRE (RADV NAVI31) AMD Ryzen 7 7800X3D 8-Core Processor 33 GB 6.17.1-cachyos1.lto.fc44.x86_64 performance
scx_bpfland | 2 Steam Runtime 3 (sniper) AMD Radeon RX 7900 GRE (RADV NAVI31) AMD Ryzen 7 7800X3D 8-Core Processor 33 GB 6.17.1-cachyos1.lto.fc44.x86_64 performance
scx_bpfland -S | 2 Steam Runtime 3 (sniper) AMD Radeon RX 7900 GRE (RADV NAVI31) AMD Ryzen 7 7800X3D 8-Core Processor 33 GB 6.17.1-cachyos1.lto.fc44.x86_64 performance
# Top runs: * **Highest FPS**: `scx_bpfland -S | 1` with a mean FPS slightly higher compared to other runs, indicating optimal performance in terms of raw frame rates. * **Smoothest FPS**: `scx_bpfland | 2` with the lowest standard deviation and variance, ensuring the most consistent frame delivery. * **Best overall**: `scx_bpfland | 2`, maintaining a good balance between performance and stability with slightly better FPS smoothness than runs with higher mean FPS but significantly worse standard deviation. # Issues: * The run with labeling as `scx_bpfland -S | 2` showed significantly higher FPS variance and lowest mean FPS, indicating stability issues compared to other rounds despite identical hardware and software configurations. # Summary This benchmark evaluation compares the Linux kernel scheduler `scx_bpfland` and its variant `scx_bpfland -S` using the same hardware setup, revealing differences in FPS stability and mean performance. The `scx_bpfland` scheduler consistently shows better frame time stability and overall performance compared to `scx_bpfland -S`, with the latter demonstrating increased variance. This analysis suggests that `scx_bpfland` is better suited for environments where frame time stability is critical, while the `scx_bpfland -S` variant may not be ideal for this particular use case, as it yields less consistent results.