PC Bottleneck Calculator (CPU vs GPU)
Estimate whether your CPU or GPU is the limiting factor for gaming, streaming, or rendering. Adjusts for resolution and task, then recommends which component to upgrade.
TECHPick your CPU and GPU, choose your target resolution and primary task, and the calculator estimates the percentage bottleneck and tells you which component is holding back the other. Useful before buying a new GPU, planning a CPU upgrade, or deciding whether a prebuilt PC is well balanced.
The tool uses relative performance scores for ~30 popular CPUs (Intel 10th-14th gen, AMD Ryzen 5000/7000/9000 including X3D) and ~30 popular GPUs (NVIDIA RTX 30/40/50, AMD RX 6000/7000). The GPU score is scaled by resolution (1080p × 0.85, 1440p × 1.0, 4K × 1.25) because higher resolutions push more pixels and lean harder on the GPU. The CPU score is scaled by task (gaming × 1.0, streaming × 1.2, rendering × 1.3) because encoding and creator workloads load the CPU more heavily. The calculator compares effective scores: balance % = abs(cpu - gpu) / max(cpu, gpu) × 100. Under 5% counts as well balanced; 5-15% is minor; 15-30% is noticeable; over 30% is severe. Worked example: an i5-13600K paired with an RTX 4070 at 1440p gaming lands at roughly 3% balance - effectively no bottleneck. Drop the same combo to 1080p and the GPU score shrinks (less pixels to render), pushing the system toward a small CPU bottleneck at high refresh rates. Crank it to 4K and the GPU becomes the clear limit while the CPU coasts.
PC Bottleneck Calculator (CPU vs GPU)
Estimate whether your CPU or GPU will be the limiting factor for gaming, streaming, or rendering. Adjusts for resolution and workload, then suggests which component to upgrade.
How a CPU/GPU Bottleneck Works
In every rendered frame, the CPU prepares draw calls, game logic, AI, and physics, then hands the work off to the GPU, which actually rasterizes pixels. Whichever chip finishes its part later sets the frame rate. If the CPU is slower than the GPU at delivering frames, you have a CPU bottleneck (GPU sits idle waiting). If the GPU is slower, you have a GPU bottleneck (CPU sits idle).
Resolution shifts the balance dramatically. At 1080p the GPU draws fewer pixels per frame, so it finishes fast and the CPU often becomes the limit. At 4K the GPU has 4x more pixels to shade, so it works far harder while the CPU's per-frame job stays the same. That is why a high-end CPU paired with a midrange GPU looks "fine" at 4K but reveals a GPU bottleneck, and the same pair at 1080p flips into CPU-limited territory.
A small bottleneck (under 10%) is normal and even healthy. You want some headroom on the non-limiting component so frame times stay consistent during spikes. Ignore the myth that "100% CPU or GPU usage = bad bottleneck." 99% GPU usage at high frame rates is actually the ideal state for gaming - the GPU is the one that should be saturated. A real bottleneck shows up as one chip pegged while the other sits well below 70%.
Approximate estimate based on relative performance scores. Real-world bottleneck also depends on RAM speed and capacity, PCIe generation, storage, GPU drivers, game engine, and specific in-game settings (ray tracing, draw distance, etc).