What a fill-rate score says about drawing large surfaces

A fill-rate score can look like a direct measure of how quickly your graphics card draws a whole screen. It is useful for comparing graphics performance, but it does not tell you exactly how a game or design app will run.

The Graphics Test checks shading power and fill rate at 1080p. Here is what the fill-rate result can tell you about drawing large surfaces, and what it cannot prove.

Run the fill-rate test

1. Open the Graphics Test on the computer you want to check.
2. Close games, video editors, and other graphics-heavy apps so they do not compete for the GPU.
3. Run the test and note the fill-rate score, along with the shading score.
4. If you want to compare results, run the test again under the same conditions. Use the same computer, browser, display setup, and power mode.

A browser graphics test is a workload, not a complete examination of the graphics card. A steady result is more useful than one run made while the computer is busy or hot.

What the result means

A higher fill-rate score

A higher score suggests the browser completed this test's pixel-focused workload more quickly. Fill rate is a rough measure of how much pixel output a GPU can process over time. A pixel is one tiny colored point on the screen, and a large surface may require the GPU to produce many of them.

This can matter for full-screen effects, large backgrounds, translucent layers, and scenes that cover much of the display. It is one part of graphics performance, not a promise that every large object will render quickly.

A lower score than expected

The score may indicate that this particular workload is slower on your device, but it does not identify the cause. The browser may be using a different graphics path, another app may be using the GPU, or the computer may be limiting performance to save power or control heat.

Compare the result with another run on the same device before drawing conclusions. Scores from different browsers, operating systems, or devices are not always directly comparable.

Fill rate and shading scores differ

That difference is normal. Fill rate focuses more on producing pixel output, while shading measures work on the calculations that determine how pixels look. A scene can be limited by either kind of work, or by something else such as memory bandwidth, texture handling, or the CPU preparing the scene.

The score is strong, but an app still feels slow

The app may be waiting on a different part of the graphics pipeline, which is the series of steps used to turn a scene into an image. It could also be limited by the CPU, storage, app settings, or the number of objects being drawn. A fill-rate result alone cannot diagnose those causes.

How to get a useful comparison

1. Close competing graphics apps. Quit games, video playback, 3D tools, and other browser tabs with animated content. Then repeat the test and see whether the score changes.
2. Use the same power conditions. On Windows 11, check Settings › System › Power & battery › Power mode. On macOS, check System Settings › Battery and compare runs with the same power setting. A laptop may reduce performance while on battery.
3. Check browser graphics acceleration. In Chrome, open Settings › System and check Use graphics acceleration when available. In Edge, open Settings › System and performance and check Use graphics acceleration when available. If you change it, restart the browser before testing again. The exact setting can vary by browser.
4. Let the computer cool down. A laptop or compact PC can reduce its speed when it gets hot. Wait a few minutes after a demanding task, keep air vents clear, and run one test again. Do not repeat heavy tests continuously just to chase a higher score.
5. Update the browser and graphics driver. If performance changed after an update or graphics acceleration is unavailable, install normal system and driver updates, restart, and test once more. Avoid changing several settings at once, so you can tell which change mattered.

If you need to check whether the browser is using GPU acceleration, see how to check if your browser uses GPU acceleration.

Why large surfaces do not always need high fill rate

The size of a surface is only part of the work. A simple, flat full-screen color can be easy to draw. A smaller surface with detailed lighting, transparency, reflections, or complex shading can take more work per pixel. Some effects also draw over the same area several times, a process called overdraw.

Resolution matters too. A 3840 by 2160 image contains four times as many pixels as a 1920 by 1080 image. That does not mean a real app will run exactly four times slower at the higher resolution. It may use upscaling, draw fewer effects, or hit another limit first. The test's 1080p result is a reference for its workload, not a direct measurement at every screen resolution.

Games also use frame rate, or frames per second, to describe how often a new image appears. Fill rate can affect frame rate when pixel output is the bottleneck, but the score does not predict FPS by itself. For example, a game with complex physics may be limited by the CPU, while a game with heavy full-screen effects may put more pressure on the GPU.

A graphics score cannot confirm that every part of a GPU or its video memory is healthy. It measures performance in a browser workload, not every possible fault. For more on that limitation, read what a browser graphics score cannot tell you about VRAM.