# Top runs: * **Highest FPS**: `scx_gamer_ltofat_cpunative_200_2000OC3` - This run achieved the highest average FPS, outperforming all others by a significant margin. * **Smoothest FPS**: `scx_gamer_1115B1` - With the lowest standard deviation and variance in FPS, this run provided the most stable frame rate throughout the test. * **Best overall**: `scx_gamer_1115B1` - While not the highest in FPS, it has excellent stability and is among the top for overall performance, making it the best balanced choice. # Issues: * The lto fat with cpu native and bugfixes run had lower GPU memory clock values and higher variance, indicating a potential issue affecting performance consistency. * The second lto fat with cpu native run with the ringbuffer had a significantly lower GPU clock and memory clock compared to the rest, potentially impacting performance negatively. * One of the lto fat with cpu native and bugfixes configurations encountered inconsistent GPU temperatures, which might affect the results' reliability. # Summary In the A vs B testing for Splitgate 2, both configurations, `LTO FAT CPU Native` and `lto fat + cpu native + ringbuffer changes`, were analyzed with all runs operating on identical hardware and operating systems. The configuration without ringbuffer changes demonstrated superior performance, especially in terms of average FPS and consistency, with `scx_gamer_ltofat_cpunative_200_2000OC3` achieving the highest FPS. However, the ringbuffer adjustments in some runs led to variations in clock speed and GPU temperature, causing those runs to be less reliable. Overall, while ringbuffer changes offer potential benefits, the current implementation underperforms compared to the baseline `LTO FAT CPU Native` setup.