Benchmark #1399

Download
Elden Ring | BORE vs SCX

CPU Limited scene in Gatefront Ruins (AI heavy) with no background load. 60 second frame recording with 0ms interval. Performance mode used in all schedulers, Performance Governor/EPP

Submitted 2 weeks ago by summedkibbles36

Specifications
Label OS GPU CPU RAM OS specific
BORE Freedesktop SDK 24.08 (Flatpak runtime) AMD Radeon RX 7900 GRE (RADV NAVI31) AMD Ryzen 7 7800X3D 8-Core Processor 33 GB 6.14.4-cachyos2.lto.fc43.x86_64 performance
BPFLAND Freedesktop SDK 24.08 (Flatpak runtime) AMD Radeon RX 7900 GRE (RADV NAVI31) AMD Ryzen 7 7800X3D 8-Core Processor 33 GB 6.14.4-cachyos2.lto.fc43.x86_64 performance
BPFLAND Next Freedesktop SDK 24.08 (Flatpak runtime) AMD Radeon RX 7900 GRE (RADV NAVI31) AMD Ryzen 7 7800X3D 8-Core Processor 33 GB 6.14.4-cachyos2.lto.fc43.x86_64 performance
BPFLAND Next --slice-us 3000 Freedesktop SDK 24.08 (Flatpak runtime) AMD Radeon RX 7900 GRE (RADV NAVI31) AMD Ryzen 7 7800X3D 8-Core Processor 33 GB 6.14.4-cachyos2.lto.fc43.x86_64 performance
BPFLAND --slice-us 3000 Freedesktop SDK 24.08 (Flatpak runtime) AMD Radeon RX 7900 GRE (RADV NAVI31) AMD Ryzen 7 7800X3D 8-Core Processor 33 GB 6.14.4-cachyos2.lto.fc43.x86_64 performance
LAVD Freedesktop SDK 24.08 (Flatpak runtime) AMD Radeon RX 7900 GRE (RADV NAVI31) AMD Ryzen 7 7800X3D 8-Core Processor 33 GB 6.14.4-cachyos2.lto.fc43.x86_64 performance
LAVD No Fork Penalty Freedesktop SDK 24.08 (Flatpak runtime) AMD Radeon RX 7900 GRE (RADV NAVI31) AMD Ryzen 7 7800X3D 8-Core Processor 33 GB 6.14.4-cachyos2.lto.fc43.x86_64 performance
LAVD No Fork Penalty --slice-max-us 1000 Freedesktop SDK 24.08 (Flatpak runtime) AMD Radeon RX 7900 GRE (RADV NAVI31) AMD Ryzen 7 7800X3D 8-Core Processor 33 GB 6.14.4-cachyos2.lto.fc43.x86_64 performance
LAVD --slice-max-us 1000 Freedesktop SDK 24.08 (Flatpak runtime) AMD Radeon RX 7900 GRE (RADV NAVI31) AMD Ryzen 7 7800X3D 8-Core Processor 33 GB 6.14.4-cachyos2.lto.fc43.x86_64 performance
# Top runs: * **Highest FPS**: `BPFLAND --slice-us 3000` with a higher mean FPS than other runs, offering a performance edge in this configuration. * **Smoothest FPS**: `BPFLAND Next --slice-us 3000` with the lowest standard deviation and variance in FPS, indicating excellent stability during the run. * **Best overall**: `BPFLAND --slice-us 3000` achieves the top balance by providing high FPS alongside a low standard deviation compared to others. # Issues: * The `LAVD` configuration exhibits lower FPS performance compared to other schedulers. The mean FPS in this setup is noticeably below the others, potentially impacting fluidity. # Summary This benchmark compared various scheduler implementations tested on a consistent hardware setup running the Freedesktop SDK 24.08 in a CPU-limited scenario in Elden Ring. Different scheduler strategies were tested for frame recording over 60 seconds, including `BORE`, `BPFLAND`, `BPFLAND Next`, `LAVD`, and their variants with slicing adjustments. `BPFLAND --slice-us 3000` emerged as the top performer, showcasing superior frame rates and stability, whereas `LAVD` lagged behind in FPS performance, suggesting that adjustments in slicing across different schedulers can have a significant impact on gaming performance and stability in CPU-heavy environments.