What Is 99th Percentile FPS? 1% Lows Explained

99th percentile FPS, usually shown as “1% lows,” is a measure of how smooth a game feels at its worst moments. It describes the frame rate you get when you look only at the slowest one percent of frames, which is where stutter hides. A game with a high average FPS but poor 1% lows can feel choppy, while one with a lower average and tight lows often feels better.

Average FPS tells you how fast a game runs overall, and it hides short slowdowns. This guide explains the difference between averages and percentiles, how frame times connect to both, why stutter shows up in the lows, how overlays and capture tools report the figure, and how to interpret the numbers without overreacting.

Average FPS Versus Percentile Frame Rates

Average FPS is the total number of frames divided by the time taken. It is a useful headline number, but it blends smooth and rough moments together. If a game runs at a steady pace for most of a minute and then freezes briefly, the average barely moves, even though the freeze is obvious to the player.

Percentile metrics fix this by sorting the frames. For the 99th percentile, you rank all frame times from fastest to slowest and look at the value that 99 percent of frames beat. Only one percent of frames are slower. That slower tail is where hitches and stutters live, so the metric captures the experience of rough moments rather than the typical one.

The phrase “1% low FPS” and “99th percentile FPS” are related and often used loosely as the same thing, though tools differ in the exact method they use. That is worth keeping in mind when you compare numbers from different sources.

Frame Time: The Raw Material

Everything starts with frame time, the number of milliseconds a single frame takes to draw. FPS is simply the inverse of this. The table below shows the relationship for a few common frame rates.

If most frames take 8 ms and a handful take 40 ms, the average frame rate stays high, but those 40 ms frames are individual stutters. The 1% low turns that slow tail into a single number you can compare.

Why Stutter Shows Up in the 1% Lows

Stutter is a sudden gap between frames. The eye and hand notice rhythm: smooth motion needs frames to arrive at an even pace. A single long frame interrupts that rhythm. The 1% low reflects the long frames directly, while the average dilutes them across thousands of normal frames.

Common causes of low 1% lows

• CPU limits: the processor cannot prepare frames fast enough in busy scenes, often showing up as spikes in crowded areas.
• Memory pressure: running out of video memory forces data to move through system RAM, creating hitches.
• Asset streaming: loading textures or world data during play can briefly delay frames, especially on slower drives.
• Background tasks: antivirus scans, updaters and overlays can steal CPU time.
• Shader compilation: some games compile shaders on the fly, which causes one-off hitches.

Why a high average can still feel bad

A game can show a high average while 1% lows sit far below it. The gap between the two is the useful signal: a wide gap suggests uneven delivery, while a narrow gap suggests consistent pacing.

How Monitoring Tools Report It

Several free and built-in tools display 1% lows. They include the performance overlay in the NVIDIA app, third-party frame-time capture utilities and the in-game benchmarks that some titles include. Each tool follows its own definition, so treat the numbers as comparable only within the same tool.

Different methods, different numbers

Some tools take the average of the slowest one percent of frames and convert it to FPS. Others take the 99th percentile frame time and convert that single value. The first method is generally a little lower than the second because it averages the worst frames rather than picking a cutoff. Both are valid, but they are not interchangeable.

0.1% lows

Some tools also show 0.1% lows, which look at an even smaller slice of the slowest frames. These catch rare, severe hitches that the 1% figure can smooth over, and they are often a closer match to the rare freezes players notice.

Sample length matters

A short capture contains few frames, so the slowest one percent is a small sample and may be dominated by one event. Longer runs, repeated several times in the same scene, give steadier results.

How to Interpret the Numbers

There is no single magic threshold, since taste and game type matter, but some general rules help.

• Compare 1% lows to the average. A modest gap usually feels smooth. A large gap suggests stutter worth investigating.
• Look at the lows relative to your monitor’s refresh rate. Lows well under your refresh rate can cause visible judder on a fixed refresh display, while a variable refresh monitor hides some of the impact.
• Compare like with like. Same scene, same settings, same tool, and same capture length.
• Remember that frame pacing matters beyond a single number. A graph of frame times shows spikes that a summary figure cannot.

Many reviewers and players treat lows near or above 60 FPS as a comfortable target for ordinary play, with higher targets for fast competitive games. That is general guidance, not a rule.

Improving Poor 1% Lows

If lows are weak, a few safe, reversible steps are worth trying before spending money on hardware.

1. Update the graphics driver and the game to their current versions.
2. Close background applications and overlays you do not need.
3. Lower the texture quality one step if video memory is nearly full.
4. Cap the frame rate slightly below what your system can sustain, which gives even pacing.
5. Install games on a fast solid-state drive to reduce streaming hitches.
6. Check whether the processor is the bottleneck by watching its usage during stutters.

A Worked Example of the Difference

Imagine two systems that both report the same average of 100 FPS in a short run. On the first system, nearly every frame takes about 10 milliseconds, with a handful stretching to 14 milliseconds. On the second, most frames take only 8 milliseconds, but a few dozen take 40 milliseconds or more. The averages match, but the second system will feel noticeably rougher because those long frames break the rhythm.

The 1% low captures this immediately. The first system’s lows sit only slightly below its average, while the second system’s lows drop sharply. A chart of frame time over the run would show a flat line for the first and a spiky line for the second.

This is why many hardware reviewers now publish both average FPS and 1% lows. The pair of numbers reveals both how fast and how consistent a system is. Treat the figures as illustrations of the method, not as measurements of any real product.

Common Mistakes When Reading Percentile Numbers

• Comparing across tools: one overlay may average the worst frames while another picks a single cutoff, so the same session can give different lows.
• Using tiny samples: a ten-second capture holds few frames and can be dominated by a single hitch or by a scene change.
• Ignoring the scene: a crowded city and an empty corridor stress the system differently, so compare the same location.
• Chasing a single number: lows are a summary. The frame-time graph tells you when and why spikes occur.
• Forgetting limits you set yourself: a frame cap, V-Sync or variable refresh can change the apparent lows, because the game paces frames on purpose.

Keeping these points in mind makes the number far more useful, and prevents over-reading small differences between runs.

Related guides

• What Does Manage Gpu Performance Counters Mean On Nvidia Hardware
• Low Profile Nvidia Graphics Cards For Small Form Factor Pcs
• Nvidia App Error Your System Configuration Is Not Supported

Frequently Asked Questions

What does 1% low FPS mean?

It is a figure that represents the frame rate during the slowest one percent of frames in a capture. It shows how rough the worst moments feel, which an average cannot reveal.

Is 99th percentile FPS the same as 1% lows?

They are closely related and often used as synonyms. Tools differ in their method, so the numbers are best compared only within the same tool.

Why are my 1% lows so much lower than my average?

A big gap points to uneven frame delivery, often from CPU limits, memory pressure, streaming or background tasks. Looking at a frame-time graph can show when the spikes occur.

Which matters more, average FPS or 1% lows?

Both matter, but lows describe smoothness better. Two systems with the same average can feel very different if one has tighter frame pacing.