Methodology
FPSBuddy uses a small set of normalized vectors to estimate FPS and detect alignment between CPU and GPU under realistic optimized settings.
1. Normalized hardware index
Every CPU and GPU is scored on a 0–100 scale relative to the current flagship (Ryzen 7 9800X3D and RTX 5090). Scores reflect average gaming performance across contemporary titles at their most representative resolution.
2. Resolution scaling
Higher resolutions shift work toward the GPU. FPSBuddy applies two coefficients per resolution — one that scales base FPS, and one that weights CPU-side bottleneck impact.
FPS = G_base × (P_gpu / 65) × R_coeff × (1 − Bottleneck% / 150)
3. Bottleneck ratio
The core ratio compares CPU vs GPU performance. Values under 1.0 indicate a CPU-bound scenario; values above 1.0 indicate a GPU-bound scenario. The ratio is then attenuated by resolution weight and per-game CPU dependence.
R = P_cpu / P_gpu → Bottleneck% = clamp(0, 65, (1 − R) × 100 × weight)
4. Upgrade recommendation
When a meaningful gap is detected, FPSBuddy scans the hardware library for the cheapest part in the opposite category whose score would push projected bottleneck under 8%.
Assumptions & limitations
Estimates assume optimized settings, no CPU thermal throttling, matched memory bandwidth, and reference driver behavior. Real-world variance of ±10–15% is normal.