Benchmark #1019

Download
Cyberpunk falcond

Submitted 1 month ago by sacredsin

Specifications
Label OS GPU CPU RAM OS specific
Cyberpunk2077_falcond_bpfland Steam Runtime 3 (sniper) AMD Radeon RX 6600 XT (RADV NAVI23) AMD Ryzen 7 3700X 8-Core Processor 34 GB 6.12.9-207.nobara.fc41.x86_64 performance
Cyberpunk2077_falcond_flash Steam Runtime 3 (sniper) AMD Radeon RX 6600 XT (RADV NAVI23) AMD Ryzen 7 3700X 8-Core Processor 34 GB 6.12.9-207.nobara.fc41.x86_64 powersave
Cyberpunk2077_falcond_lavd Steam Runtime 3 (sniper) AMD Radeon RX 6600 XT (RADV NAVI23) AMD Ryzen 7 3700X 8-Core Processor 34 GB 6.12.9-207.nobara.fc41.x86_64 powersave
Cyberpunk2077_falcond_rusty Steam Runtime 3 (sniper) AMD Radeon RX 6600 XT (RADV NAVI23) AMD Ryzen 7 3700X 8-Core Processor 34 GB 6.12.9-207.nobara.fc41.x86_64 performance
Cyberpunk2077_gamemode Steam Runtime 3 (sniper) AMD Radeon RX 6600 XT (RADV NAVI23) AMD Ryzen 7 3700X 8-Core Processor 34 GB 6.12.9-207.nobara.fc41.x86_64 performance
# Top runs: * **Highest FPS**: `Cyberpunk2077_falcond_flash` with approximately 4% higher average FPS than others, indicating excellent performance under the `powersave` scheduler. * **Smoothest FPS**: `Cyberpunk2077_falcond_bpfland` exhibited the lowest standard deviation in FPS, making it the most stable run. * **Best overall**: `Cyberpunk2077_falcond_flash` offers the best combination of highest FPS and acceptable stability, making it a top-performing run. # Issues: * The run utilizing the `lavd` scheduler had lower FPS and a higher variability than others and thus is notably underperforming compared to the rest. * The run using the `rusty` scheduler also exhibited lower FPS and higher variance, further supporting its unsuitability for optimal gaming performance. # Summary This benchmark compares various scheduling configurations for the Linux kernel while running Cyberpunk 2077. Among the tested configurations, the `powersave` scheduler provided the best FPS performance, showcasing around a 4% improvement over the others, while maintaining acceptable stability. The `performance` schedulers showed varying results, with one version miscued by the stable `bpfland` run, delivering smoother stability than most others. In contrast, the `lavd` and `rusty` configurations underdelivered in performance metrics, showing the potential trade-offs associated with different schedulers in terms of both performance and consistency.