My Kingston Fury throughput is all over the place in Cyberpunk; should I quantify the bottleneck?

I ran a scenario simulation for high-load exploration in Night City and realized that controller load peaks jumping every 0.3-0.5 seconds were causing those throughput spikes. I tried messing with the disk queue depth, and while the raw read/write speeds went up, the overall stability was still garbage. The conclusion was clear: I had to go into the BIOS and enable Re-Size BAR (Resizable BAR), then optimize the power strategy to ensure the render benchmark didn't choke. The sequence was: BIOS -> Re-Size BAR -> 0.3-0.5s peaks -> Stable Curve. Tuning performance benchmarks like this is all about predicting timing conflicts between the CPU and RAM. I could actually feel a bit of heat radiating from the memory heatspreaders during high-frequency bursts, and the keyboard response felt slightly mushy during frame drops. After validating with 3DMark, the bottleneck quantification was successfully exported, and the results are finally reliable. This is the only way to get a real read on the hardware.