Averages hide the moments that matter
Every frame rate figure you see quoted, including ours, is an average. Averages are a reasonable way to compare hardware and a poor way to describe how a game feels.
What you actually notice is the bottom of the distribution — the 1% lows, the worst one frame in a hundred. A game averaging 90 fps with lows of 80 feels smooth. A game averaging 90 fps with lows of 35 feels broken, and it is the second one that makes people think their PC is faulty. The dips arrive exactly when the scene gets busy, which is exactly when you are least able to tolerate them.
That is why this tool gives you two answers rather than one. The minimum column clears your target on average and will drop below it regularly. The comfortable column aims about 15% higher, which is roughly what it takes to keep the lows near the number you actually wanted.
What each refresh rate is worth
The returns are not linear, and it is worth knowing where they fall off before spending.
- 60 Hz is the baseline. Perfectly playable for most single-player games, and still where the majority of monitors in use sit.
- 120 to 144 Hz is the big jump. Almost everyone notices it immediately, and not only in games — dragging a window or scrolling a page feels different. If you upgrade one thing about your setup, this is usually it.
- 165 to 240 Hz is a real but much smaller improvement, and it costs considerably more hardware to sustain. It matters most in competitive shooters where reaction time is the point.
- 360 Hz and beyond is for a narrow group of competitive players, and the hardware needed to feed it is out of proportion to what most people gain.
For most builds, a 1440p 144 Hz panel is a better use of money than a 1080p 240 Hz one. You will see the extra resolution constantly and the extra refresh rate rarely.
Resolution picks your card, target picks your chip
Change the resolution above and watch which recommendation moves. The graphics card changes completely. The processor barely moves at all.
This is not a quirk of our model, it is how rendering works. Rendering more pixels is more work for the graphics card and no extra work for the processor, which is still issuing the same draw calls, running the same physics and the same game logic regardless of how many pixels the result is drawn into.
The practical consequence: if you are chasing a high refresh rate at 1080p you need a strong processor and a moderate card. If you are chasing 4K at 60 you need a strong card and a moderate processor. Those are almost opposite shopping lists, and buying the wrong one is the most common expensive mistake in PC building.
Variable refresh rate changes the calculation
If your monitor supports G-Sync or FreeSync, the target matters less than it used to. Without it, a frame rate that does not divide evenly into your refresh rate causes tearing or, with V-Sync on, sudden drops to half rate. With it, the monitor simply matches whatever the graphics card produces, so 95 fps on a 144 Hz panel looks smooth rather than uneven.
In practice this means you do not need to hit your panel's full refresh rate to have a good experience. Comfortably exceeding two-thirds of it, with stable frame times, is a perfectly sensible goal and considerably cheaper.
What this tool does not account for
Upscaling. DLSS, FSR and XeSS render at a lower internal resolution and reconstruct the image. A card running 4K with quality upscaling behaves closer to one running 1440p, which can move you a whole tier down the recommendation. Our numbers are native-resolution estimates; treat upscaling as headroom on top.
Frame generation. This raises the displayed frame rate without lowering input latency the way real frames do. It is genuinely useful for smoothness and genuinely not the same thing, so we do not fold it into an average.
Ray tracing. Turning it on can halve frame rate. Our estimates assume rasterised rendering at the stated preset.
If you want to check how a specific pairing you already own performs rather than what to buy, the bottleneck calculator works forwards from your hardware.
Common questions
Do I actually need 144 Hz?
The jump from 60 to 120 is large and most people notice it immediately, in cursor movement and camera panning as much as in games. From 144 to 240 the returns shrink sharply and mostly matter in competitive shooters. If you are choosing where to spend, a 1440p 144 Hz panel is a better buy for most people than a 1080p 240 Hz one.
Why is my frame rate lower than this tool says?
Our estimate is an average across typical play. Dense scenes, large open areas and heavy effects all run below the average. Background applications, an older driver, slow memory or a processor that cannot hold its boost clocks under heat will all cost you frames the model cannot see.
Why does the processor recommendation barely change with resolution?
Because rendering more pixels gives the graphics card more work and gives the processor none. Your resolution decides which card you need; your frame rate target decides which processor you need. They are almost independent choices.
Does DLSS or FSR change these numbers?
Yes, substantially, and we do not model it. Upscaling renders at a lower internal resolution and reconstructs the image, so a card running 4K with quality upscaling behaves closer to one running 1440p. Treat our figures as native-resolution estimates and consider upscaling headroom on top.