FPS Calculator

It starts from the graphics card’s 3D benchmark score, because the card is the primary driver of frame rate, then applies the processor as a secondary factor that caps the result when it is much weaker than the card. On top of that it applies multipliers for resolution, graphics preset, how demanding the specific game is, and memory capacity. The output is an average frame rate plus 1% and 0.1% low figures, which describe consistency rather than peak speed.

It gives you the right order of magnitude and the right direction, not a number you should hold us to. Real frame rates vary with the scene you are standing in, the driver version, the game patch, your memory speed, thermal behaviour and what else is running. Use it to compare options — this card against that one, 1440p against 4K, high against ultra — which is what it is genuinely good at. For an exact figure in one game, benchmark that game.

They describe your worst frames rather than your typical ones. The 1% low is roughly the frame rate during the worst 1% of frames, and the 0.1% low the worst one in a thousand. They matter because smoothness is what you perceive, not the average: 90 FPS average with a 45 FPS 1% low feels visibly worse than a steady 70 FPS. A wide gap between average and lows is the signature of stutter, usually from a processor limit, memory pressure or texture streaming.

It depends on what you play. 60 FPS is the comfortable baseline for single-player games. Competitive shooters benefit from 144 FPS or more, which is why this calculator judges those titles against a higher target. At 4K the expectation drops, so the target is scaled down accordingly. Above your monitor’s refresh rate the extra frames only reduce input latency — matching your display is usually the more sensible goal than chasing the biggest number.

More than most people expect, for very little visible difference. Moving from ultra to high typically returns around 25% on this model, and ultra settings are the classic case of large cost for small gain. Ray tracing is the extreme example, roughly halving the frame rate. The results panel lists what each resolution and preset would give for your exact hardware, so you can see the trade rather than guess at it.

Only up to the point where you have enough. Going from 8 GB to 16 GB usually helps, because 8 GB forces the system to page to disk in modern titles, especially above 1080p. Beyond 16 GB the gain is small — this model applies a modest uplift for 32 GB or more, which reflects reality. If you are short on memory the symptom is stutter and long hitches rather than a low average, so adding memory fixes smoothness more than it raises the headline number.

Not directly, because the gain depends heavily on which upscaler, which quality mode and which game. As a rough guide, quality-mode upscaling lands somewhere near the frame rate of the next resolution down — so 4K with quality upscaling behaves roughly like native 1440p. You can approximate it here by estimating at the lower resolution. Frame generation is a different thing again and multiplies the displayed frame rate without reducing input latency in the same way.

It currently covers 15 named titles: Cyberpunk 2077, Call of Duty: Modern Warfare III, Fortnite, Apex Legends, Elden Ring, Starfield, Valorant, Grand Theft Auto V, Minecraft, Counter-Strike 2, Red Dead Redemption 2, The Witcher 3, Rainbow Six Siege, Call of Duty: Warzone and Forza Horizon 5. If your game isn’t on the list, the closest approximation is to reason about which bucket it resembles: a lightweight competitive shooter behaves more like Valorant or CS2, and a heavy open-world game behaves more like Cyberpunk 2077 or Starfield. Everything else uses the baseline scaling.

The calculator falls back to a fixed benchmark score rather than refusing to produce a number or inventing one that looks precise. That fallback is deliberately conservative, so a missing part will generally understate rather than overstate performance. If your part is genuinely missing from the database, that’s worth reporting so it can be added with a real score.

Because the underlying relationship between benchmark score and playable frame rate stops being reliable at extreme values, and very high frame rates are increasingly limited by the game engine, the CPU, or the monitor rather than the card. The cap keeps the top end of the estimate realistic instead of extrapolating a straight line into numbers no game engine actually sustains.

A GPU bottleneck means the card is the limiting part: raising resolution or settings lowers your frame rate noticeably, because the card is doing more work per frame. A CPU bottleneck means the processor is the limiting part: raising resolution barely changes your frame rate, because the card has headroom the processor isn’t feeding it fast enough to use. In this calculator that shows up directly as the CPU factor being lower than the GPU factor for your chosen hardware. For an exact percentage rather than an implied one, use this site’s dedicated bottleneck calculator with your resolution set correctly.

Because 4K is roughly four times the pixel count of 1080p, so the same hardware produces meaningfully fewer frames there, and expecting an unchanged target is how people end up disappointed by expensive hardware. This calculator reduces the target by 30% at 4K so the rating reflects what’s actually realistic at that resolution, not the 1080p bar applied to a much harder workload.

On this model, 1440p returns about 72% of the 1080p figure for the same hardware, settings and game, a roughly 28% drop. 4K returns about 45% of the 1080p figure, a much steeper cost, which is why the resolution comparison shown alongside your result is usually more informative than the headline number on its own.

That depends entirely on whether the visual difference matters to you more than the frame rate, and this calculator won’t tell you which side of that trade you’re on, only how large it is. Ray tracing is modeled as roughly halving the frame rate compared with high settings, the single largest preset penalty in the model. Whether that’s worth it is a personal call the numbers can inform but not make for you.

Because there’s no single agreed methodology. Different tools use different benchmark sources, different divisors, different game weightings, and some model DLSS, FSR or frame generation while others don’t. None of them are measuring your PC; they’re all approximating from published data using their own formula. That’s exactly why this page documents its formula rather than asking you to trust a number with no visible working.

For comparing options, yes, that’s what it’s built for: this card against that one, 1440p against 4K, high against ultra, all using the same consistent method so the comparison is fair. For an exact frame rate in one specific game before a purchase decision that really matters to you, no estimate from any calculator replaces a real benchmark of that game, because engine-specific behavior, scene complexity and driver optimization aren’t captured by a benchmark-score formula.

No, storage isn’t a factor in this model. In reality, a slow drive can cause stutter as textures stream in, which shows up as poor 1% and 0.1% lows despite a reasonable average, but that’s a real-world effect this calculator doesn’t attempt to quantify. If your measured lows are consistently worse than estimated and you’re running from a hard drive or a heavily-fragmented SSD, storage is a reasonable suspect.

Not directly. The calculator works from each part’s stock benchmark score, so it reflects out-of-the-box performance. An overclock that meaningfully raises a part’s real-world benchmark score would push the actual result above the estimate, but the calculator has no way to know your specific overclock, so it isn’t modeled.

The average is the typical frame rate across a session. The 1% low and 0.1% low describe your worst frames, calculated as 70% and 50% of the average respectively. The frame time figures convert those same numbers into milliseconds per frame, which is how frame pacing tools like RTSS or Frameview actually measure and display consistency, useful if you’re comparing against a real capture rather than another FPS number.

Because open-world engines are typically simulating far more at once, more geometry, more physics, more systems running in the background, than a level-based or arena shooter. This calculator applies a 30% penalty to titles in that category to reflect the real, consistent gap between how those engines perform and how lighter, more optimized esports engines perform on the same hardware.

Because those games are built and optimized specifically to run on a huge range of hardware at very high frame rates, which is a competitive requirement for the genre, not an accident. This calculator applies a 50% uplift to that category, reflecting the real, well-documented gap between those engines and a typical AAA title at the same settings.

No, the calculator doesn’t know your monitor’s refresh rate and uses a fixed target, 60 FPS for most games and 144 FPS for competitive titles, reduced at 4K. In practice, frames above your monitor’s refresh rate mostly reduce input latency rather than improving what you see, so matching your actual display is usually a more useful personal target than chasing the highest number the hardware can produce.

It’s a reasonable, consistent, frequently-updated proxy, which is why it’s the basis for this calculator rather than a synthetic in-house score. It measures general graphics throughput rather than any one game’s specific engine behavior, which is exactly why the game-specific multipliers exist on top of it, and why the result is presented as an estimate rather than a guaranteed number.

No, the calculator assumes a system dedicated to running the game. In practice, background capture, overlay and streaming software all consume CPU and sometimes GPU resources and will pull your real frame rate below the estimate, particularly on systems where the CPU factor is already the limiting one.

Because a single number invites a single decision, and the more useful question is usually a comparison: is dropping to 1440p worth avoiding a drop to high settings, or vice versa. Showing your exact hardware recalculated at every other resolution and preset turns the result into a decision tool rather than a single data point.

Because the effect is really about avoiding a shortfall, not adding raw power. 8GB is penalized above 1080p because that’s where modern titles are more likely to exceed it and start paging to disk, which shows up as stutter more than as a lower average. Beyond 16GB the benefit is small, and this model applies a modest, consistent uplift at 32GB or more rather than treating additional memory as a major performance lever.

No, and that is a deliberate limit rather than a missing feature. Per-game numbers require measured benchmark runs for every combination of game, processor, graphics card and settings, and tools that present them for hundreds of titles are usually interpolating from a much smaller set of real measurements. This calculator estimates from relative hardware performance instead, so treat the result as the level your hardware supports rather than a prediction for one title.

Yes, with the same caveat that applies to any FPS estimator. Mobile chips are scored separately from their desktop equivalents, so a laptop graphics chip is not treated as its desktop namesake. What cannot be seen from the model name is the power limit and cooling the manufacturer chose, and on laptops that varies widely between machines using identical silicon. Read a laptop estimate as the ceiling.

Accurate enough to compare options, not accurate enough to promise a number. Benchmark scores capture how fast parts are relative to one another, which is what you need when deciding between two upgrades. They cannot capture how a particular game engine distributes work, how well it threads, or how much memory bandwidth it wants. Expect the shape of the answer to hold and the exact figure to move.

Select your processor and graphics card above and set your resolution. The estimate reports average FPS alongside 1% lows, which matter more for how smooth gameplay feels. For checking it on your own machine, see how to check GPU usage while gaming.

They are the slowest one percent of frames. A high average with poor 1% lows feels worse than a lower but steadier average, because the dips are what you notice as stutter.

Ideally a sustained 144 FPS, though anything above roughly 100 feels substantially smoother than 60. At high refresh rates the processor matters more, which our 1080p graphics card guide explains.

Common causes are thermal throttling, memory running single channel, or XMP left disabled in the BIOS. Our guide to fixing a CPU bottleneck covers all three, and none of them cost anything.

They are different names for the same thing and people search for each. Whether you call it an FPS checker, an FPS test, an FPS predictor or a frame rate calculator, the job is to estimate the frame rate a given processor and graphics card produce at a chosen resolution. Nothing is installed and nothing runs on your machine.

It works as a CPU GPU FPS calculator across titles generally rather than for one game specifically. A game fps calculator that promises per-title numbers for hundreds of games is usually interpolating from a much smaller set of real measurements. This pc fps calculator estimates from relative hardware performance, which is honest about what it can and cannot know.

No. An FPS counter or overlay measures what your machine is doing right now, which is useful for confirming a result but not needed to get one. This estimates before you buy or install anything, which is when the number is most useful.