# Top runs: * **Highest FPS**: `Flash Next -m all -C 0 -c 0` with approximately 1% higher FPS than the other configurations. * **Smoothest FPS**: `Flash Next -m all -C 0` with the lowest standard deviation and variance of FPS among all runs. * **Best overall**: `Flash Next -m all -C 0` combines high FPS with low variability, offering a balanced performance profile in this benchmark. # Issues: * The BORE run has lower FPS and higher variance in FPS compared to all Flash Next variants, suggesting it underperforms in both speed and stability. # Summary This benchmark comparing the `BORE` system with various `Flash Next` configurations on a GPU limited scene in Elden Ring shows that the `Flash Next` configurations generally outperform the `BORE` system in terms of FPS. Notably, the `Flash Next -m all -C 0` run provides the most consistent performance with the least variability, while the `Flash Next -m all -C 0 -c 0` variant achieves the highest average FPS. Though all tested under similar conditions with the same hardware and kernel parameters, the `BORE` run falls short in both average FPS and stability, highlighting the superior performance of the `Flash Next` configurations for this specific AI-heavy scene.