Why Reviewers Measure 0.1% and 1% Low Frame Times

This article explains why computer hardware reviewers measure both 1% and 0.1% low frame times alongside average frame rates during benchmarks. While average frames per second (FPS) indicate overall performance, they fail to capture the stuttering and lag spikes that ruin a user's experience. By breaking down the math and practical differences between 1% lows and 0.1% lows, this guide illustrates how reviewers use these specific metrics to expose micro-stutters, hardware bottlenecks, and the true smoothness of a CPU or GPU.

The Problem with Average FPS

For years, average FPS was the standard metric for benchmarking PC hardware. However, averages hide inconsistencies. A system might average 100 FPS over a 60-second benchmark, which sounds smooth. Yet, if that system renders frames at 120 FPS for 58 seconds and freezes completely for 2 seconds, the average is still high, even though the actual experience is terrible.

To measure smoothness rather than just raw speed, reviewers look at "frame times"—the exact amount of time, in milliseconds, it takes for the hardware to render a single frame. The longest frame times are then converted back into an FPS figure to give us the 1% and 0.1% lows.

What 1% Lows Tell Us

To calculate the 1% low, benchmarking software records the frame time of every single frame rendered during a test, sorts them from fastest to slowest, and isolates the slowest 1% of those frames.

For example, if a benchmark renders 10,000 frames, the 1% low is the average speed of the slowest 100 frames. This metric represents the noticeable dips in performance you feel during demanding moments in a game, such as a massive explosion or entering a crowded city. If a game has a 100 FPS average and an 80 FPS 1% low, the gameplay will feel very consistent.

Why Reviewers Also Need 0.1% Lows

While 1% lows are great for showing general performance dips, they are still an average of a group of frames, meaning they can hide brief, catastrophic hitches. This is why reviewers also measure 0.1% lows.

Using the same 10,000-frame benchmark example, the 0.1% low looks at only the 10 slowest frames. This metric specifically targets micro-stutters and momentary freezes.

Here is why reviewers cannot rely on 1% lows alone:

Exposing Micro-Stutters: A game might only stutter once every ten seconds. Because those stutters represent so few frames, they might get averaged out and hidden within the 1% low data. The 0.1% low catches these isolated spikes.
Identifying CPU Bottlenecks: Modern CPUs can hit 100% utilization for a fraction of a second while loading background assets, causing a jarring hitch. This manifests immediately in the 0.1% low column, even if the 1% low looks acceptable.
Catching VRAM Limits: When a graphics card runs out of video memory (VRAM), it has to pull data from the much slower system RAM. This causes severe, split-second frame time spikes that tank the 0.1% lows long before they heavily impact the 1% lows or average FPS.
Evaluating Engine Optimization: Sometimes, poor 0.1% lows are the fault of the software, such as shader compilation stutter. Reviewers use 0.1% lows to warn consumers if a game is poorly optimized, regardless of how powerful the hardware is.

How to Read the Metrics Together

Reviewers present Average FPS, 1% lows, and 0.1% lows together to paint a complete picture of performance. When looking at a benchmark chart, the goal is to see all three numbers as close to each other as possible.

A small gap between the 1% low and the 0.1% low means the hardware is delivering a stable, fluid experience. A massive drop between the 1% low and the 0.1% low tells the reviewer—and the consumer—that despite high frame rates, the hardware is suffering from severe, immersion-breaking stutters.