How We Test Laptops

The laptop review process at PCMag.com carries on core traditions that date back to the establishment of PC Labs in 1984: We compare each system to others in its category on the basis of price, features, design, and hands-on, repeatable performance tests.

To evaluate performance, we use an array of benchmark software and real-world applications and games, carefully chosen to highlight the strengths and weaknesses of a PC's mix of components. That evaluation ranges from the CPU and the memory subsystem to the machine's storage hardware and graphics silicon. We test production units with the latest updates and drivers available; pre-production or prototype systems are described in preview stories, not reviews with benchmark results and five-star-scale ratings.

We regularly evaluate new benchmark solutions as they hit the market and overhaul our testing procedures to keep pace with the latest technologies. In August 2021, we rolled out a new suite of benchmarks, replacing those we'd been using for several years. The downside of changing our benchmarks is that it resets the database of tested PCs that we can use for comparisons in reviews (though we have been running the new tests on every recent notebook we can get our hands on to build it back up). The upside is that it gives us more current, more accurate comparisons that will only get better as we add more data.

Our laptop testing focuses on four roughly divided aspects of performance: productivity and content creation, graphics and gaming, battery life, and display quality. Here's a breakdown of each. Except for our battery rundown, we conduct tests with the laptop running on AC power and Windows 10's power profile set to best performance rather than battery conservation.

Productivity Testing

PCMark 10

Our first (and arguably most important) benchmark test is UL's PCMark 10. This wide-ranging suite simulates a variety of Windows programs to give an overall performance score for office workflows. The tasks involved include such everyday staples as word processing, web browsing, videoconferencing, and spreadsheet analysis.

We run the primary PCMark 10 test (not the Express or Extended versions), which yields a proprietary numeric score. Results over 4,000 or 5,000 points indicate excellent productivity for everyday Microsoft Office or Google Workspace tasks.

Results from the PCMark 10 test can be used to compare systems' relative performance for everyday tasks. (PCMark 10 is also used by large organizations to gauge potential new hardware's suitability to workloads, compared with existing installed hardware.) Bear in mind, however, that PCMark 10 results, like those from most of our benchmark tests here, are highly sensitive to the specific configuration of the PC being tested.

PCMark 10 Full System Drive Storage Test

We also run PCMark 10's Full System Drive storage subtest, which measures the program load time and the throughput of the laptop's boot drive. Nowadays, that is almost always a solid-state drive rather than a spinning hard drive.

Like the productivity test, the PCMark 10 Storage test delivers a numeric score with higher numbers indicating quicker response.

The benchmark is designed to factor in lower-end Serial ATA bus architectures and higher-end PCI Express/NVMe ones alike. It aims to quantify the real-world differences in performance attributable to these different drive types. (Earlier versions tended not to differentiate much between various implementations of SSD.)

Cinebench R23

Maxon's Cinebench is a CPU test that uses the company's Cinema 4D engine to render a complex scene. We use the test's multi-core benchmark, which exercises all of a processor's cores and threads—the more powerful the chip, the higher the score. Cinebench tends to scale well with more cores and threads, as well as with higher clock speeds.

Cinebench is a raw test of a PC's number-crunching ability, and it is paralleled in applications ranging from spreadsheet summations to computer-aided design and 3D rendering. The score reflects how well a laptop will handle processor-intensive workloads.

Geekbench 5.4 Pro

Primate Labs' Geekbench is another processor workout. It runs a series of CPU workloads designed to simulate real-world applications ranging from PDF rendering and speech recognition to machine learning.

We record Geekbench's Multi-Core Score. (Higher numbers are better.) Geekbench is especially handy because, unlike many other common benchmark tests, it has versions for many platforms (including Apple's macOS and iOS), which can enable valuable cross-comparisons.

HandBrake 1.4

Video editing is one of the most demanding tasks for a PC. HandBrake is an open-source video transcoder for converting multimedia files to different resolutions and formats.

We record the time HandBrake takes, rounded to the nearest minute, to encode a 12-minute 4K H.264 video file (the Blender Foundation movie Tears of Steel) to a more compact 1080p version. We use the software's Fast1080p30 preset for this conversion.

This is primarily a CPU test. Like Cinebench, it scales well with more cores and threads, and in systems that have the robust thermals to handle heavy, sustained processing loads over several minutes. Because this is a timed test, lower times are better.

PugetBench for Photoshop (Using Adobe Photoshop 22 CC)

Our final productivity test is Puget Systems' PugetBench for Photoshop, which uses version 22 of Adobe's popular image editor to measure a PC's performance for content creation and multimedia applications. PugetBench is an automated Photoshop extension that replaces our former stopwatch test, which ran through a scripted workload using a 2018 version of the software.

PugetBench executes a broad range of general and GPU-accelerated Photoshop tasks ranging from opening, resizing, rotating, and saving an image to applying masks, gradient fills, and filters (including Lens Correction, Smart Sharpen, Field Blur, and Tilt-Shift Blur).

The PugetBench for Photoshop Overall Score is a numeric value based on a 50/50 split between the general and filter tasks. (Higher numbers are better.) This benchmark exercises both the system's CPU and its graphics chip or card, as well as its memory and storage subsystems. Note that low-memory and budget systems (such as those Windows 10 models with just 4GB of RAM) may not be able to complete this test.

Graphics Testing

Whether a laptop relies on integrated graphics built into the processor or a discrete GPU like one of Nvidia's GeForce or AMD's Radeon series, its graphics capabilities affect everything from whether it can play the latest games to how promptly application windows appear on screen. We use some benchmarks that report proprietary scores and others that measure frames per second (fps), the frequency at which the graphics hardware renders frames in a sequence, which translates to how smooth the scene looks in motion.

3DMark Night Raid and 3DMark Time Spy

UL's 3DMark is a graphics test suite for Windows that contains a number of benchmarks for different GPU functions and software APIs. We run two DirectX 12 tests on all PCs: 3DMark Night Raid, and 3DMark Time Spy.

Night Raid is the more modest of the two test workloads, suitable for laptops with integrated graphics or for mobile devices. It's meant for lower-power, mainstream systems and renders at a simulated resolution of 1,920 by 1,080 pixels.

Time Spy, meanwhile, is much more demanding, suitable for high-end PCs with the latest graphics cards. It renders at a simulated resolution of 2,560 by 1,440 pixels. The test leverages features of the DX12 API, including asynchronous compute, explicit multi-adapter, and multi-threading.

Each test yields an Overall Score, which is what we report. (We don't break out the graphics and CPU scores.) Higher numbers are better.

3DMark Port Royal

We run the 3DMark Port Royal test only on certain compatible (mostly, gaming) laptops. Primarily, it is portables with Nvidia's GeForce RTX series graphics silicon that get this third 3DMark benchmark, which is a real-time ray-tracing exercise.

Ray tracing is a 3D rendering technique that traces the path of light from your viewpoint through the pixel plane and into the scene, where it may reflect from or be blocked by objects or pass through transparent or semi-transparent objects. It delivers greater visual realism at the expense of greater computational complexity.

The test renders at a simulated 2,560 by 1,440 pixels. Port Royal reports a numeric score based on how smoothly a PC presents an animated scene.

GFXBench 5.0

GFXBench is a cross-platform GPU performance benchmark that stress-tests both low-level routines like texturing and high-level, game-like image rendering. We run two of its subtests, Aztec Ruins and Car Chase, rendered off-screen to accommodate different display resolutions and make comparisons valid.

Both Aztec Ruins (1440p) and Car Chase (1080p) exercise graphics and compute shaders, but the former relies on the OpenGL application programming interface (API) while the latter uses hardware tessellation. We record the results in frames per second (fps); higher numbers are better.

Like Geekbench, GFXBench is extra-valuable as a benchmark test because it allows for multiplatform (that is, not just Windows) cross-comparisons. Beyond Windows 10, it accommodates devices using Android, Apple's macOS, and Apple's iOS.

Real-World Gaming & Content Creation Tests

The synthetic tests above are helpful for measuring general 3D graphics aptitude, but it's hard to beat full retail video games for quantifying gaming performance. To that end, we’ve chosen three games that each represent a major genre that gamers play, and that stress your hardware in different ways: a big-budget AAA title, a sports/racing sim, and a hyper-popular esports shoot-'em-up.

All three of these games include built-in benchmark tests that run through a consistent scene or snippet of gameplay and measure frame rates, delivering results in frames per second. The default test resolution is 1080p (1,920 by 1,080 pixels), but we’ll also test at a laptop’s native resolution for anecdotal results if the display is 1440p (2,560 by 1,440) or 4K (3,840 by 2,160).

Assassin's Creed Valhalla

Assassin’s Creed Valhalla is our representative for big-budget (or “AAA”) single-player games, the bread and butter for many performance-conscious gamers. These are usually action-adventure or RPG titles that test the limits of visual realism and increasingly include big open worlds. These games push cutting-edge character models, textures, lighting, and expansive landscapes that require some serious processing and graphics power.

Valhalla is a 2020 action role-playing game from Ubisoft. We run this test on the Medium and Ultra High graphics quality presets.

F1 2021

The two other games we test are F1 2021 and Tom Clancy's Rainbow Six Siege, which are less visually demanding and serve similar (but still distinct) purposes. F1 2021 is, of course, a Formula One racing sim, so high frame rates for smooth gameplay are a must.

We use this Codemasters title to test the action-sim genre. We run F1 2021's built-in bench test on the Monaco test track at 1080p and the Ultra High quality preset.

We also, in some cases, use this test to see how much Nvidia’s DLSS technology can help less powerful laptops. Our first F1 2021 test run is done with TAA anti-aliasing, but on a second run we switch the anti-aliasing setting to DLSS, if supported, to improve frame rates, and take note of the improvement.

Tom Clancy's Rainbow Six Siege

Rainbow Six Siege, another Ubisoft title, also thrives on high frame rates and represents the mega-popular competitive multiplayer/esports genre, particularly first-person shooters. This title is not meant to be visually demanding, but like the classic Counter-Strike: Global Offensive, is optimized to run at high frame rates and to scale well to a wide variety of hardware.

Gamers playing on weak laptops will be happy to see 60fps scores, but enthusiasts with powerful machines want to see frame rates in the hundreds, often to make use of their high-refresh screens that provide a competitive edge. We test this game at the Low and Ultra quality presets, with render scaling manually set to 100%.

Recommended by Our Editors

How We Test Everything We Review

The Best 17-Inch and 18-Inch Laptops for 2024

The Best Cheap Laptops for 2024

PugetBench for Premiere (Using Adobe Premiere Pro 15 CC)

In addition to testing all compatible laptops with Puget Systems' PugetBench for Adobe's Photoshop image editor, we run PugetBench for Adobe's Premiere Pro video editor on systems suitable for that challenging app: gaming and content-creation laptops.

This automated extension performs live playback and file export with several codecs at 4K and 8K resolutions, as well as special-effects sequences that stress the CPU and GPU beyond normal Premiere Pro operations. We report the test's standard overall score. Like with the PugetBench Photoshop test, higher numbers are better.

Laptop Battery Testing

How long a notebook will run without being plugged into a power outlet varies wildly depending on what it's doing—playing a game, browsing the web, writing emails, or whatever—and factors such as its screen resolution (more pixels to illuminate will drain the battery faster) and brightness setting.

We aim for a fair test of relative battery life by giving all laptops the same job: playing a locally stored 720p video file (the open-source Blender movie Tears of Steel) with display brightness at 50% and audio volume at 100% until the system quits. We make sure the battery is fully charged before the test, with Wi-Fi and keyboard backlighting turned off. We set Windows 10's power profile to the "better battery" (or equivalent) setting rather than "peak performance" and configure the system to shut down (hibernate) when battery charge has fallen to 5%.

A few laptops, mostly Chromebooks with eMMC flash storage rather than solid-state drives, don't have enough onboard storage capacity to hold our Tears of Steel video file. In those cases, we play the movie from a solid-state drive connected to a USB port.

Laptop Screen Brightness & Color Testing

To quantify the brightness output and color gamut of the laptop screen, we use Datacolor's SpyderX Elite colorimeter/calibrator tools to measure display performance.

We test screen brightness output levels at 50% and 100% brightness, per the OS's brightness setting. Results are reported in nits (candelas per square meter). We also use the Spyder tool and Datacolor's software to measure the gamut coverage (by percentage) for the three most relevant color spaces for laptop users: the sRGB, Adobe RGB, and DCI-P3 standards.

A color space is a defined range of colors. Photographers will be most interested in sRGB and Adobe RGB, as they were developed with photos and still images in mind. sRGB is the standard color space for the web and many other applications. That said, Adobe RGB does have a considerably wider color gamut than sRGB, and savvy photographers may be able to coax a more vivid range of colors when printing Adobe RGB images, whether they're outputting their own prints or working with a commercial service.

DCI-P3 is more cinema-oriented. It was defined by the Digital Cinema Initiatives (DCI), a consortium of major motion-picture producers. It is geared toward digital video projection. If you are purely a photographer, or a photo editor, you can ignore DCI-P3 results, but they will come into play if you work with video.

Special Cases: MacBooks, Chromebooks, and Mobile Workstations

Being Windows benchmarks, PCMark 10 and 3DMark don't run on macOS and Chrome OS laptops. There are, however, macOS versions of the Cinebench and Geekbench CPU tests and GFXBench GPU tests, which we do run.

The latest version of the HandBrake video encoder also supports Apple's M1 processor. At this writing in August 2021, Adobe Photoshop extensions did not work on M1 MacBooks, but the PugetBench team was working on porting their Photoshop benchmark to the platform.

Chromebooks

For Chromebooks, we run one Chrome OS benchmark suite: Principled Technologies' CrXPRT 2. It measures how quickly a system performs everyday tasks in six workloads (applying photo effects, finding faces in images, encrypting and displaying offline notes, calculating and graphing views of a stock portfolio, analyzing DNA sequences, and generating 3D shapes using WebGL). The performance test yields a numeric score; higher numbers are better.

We also run Basemark Web 3.0, which measures how well a Chromebook can run web- or browser-based applications. It combines low-level JavaScript calculations and tests using popular JavaScript frameworks and Document Object Model and CSS features with WebGL graphics content that exercises the GPU. The test also yields a numeric score; higher numbers are better.

Since Chromebooks can run Android apps, we also run the Android versions of the Geekbench CPU and GFXBench Aztec Ruins and Car Chase GPU tests, as well as the Work 3.0 benchmark from UL's PCMark for Android.

Mobile Workstation Laptops

One last class of laptops we test is the mobile workstation. Mostly running Windows 10, these powerful machines tend to run Xeon processors and Quadro GPUs, though there is some crossover with the components in high-end power laptops.

With these professional-oriented laptops, we run two additional benchmarks. The first, Blender, is an open-source 3D content creation suite for modeling, animation, simulation, and compositing. We record the time it takes for Blender's built-in Cycles path tracer to render two photo-realistic scenes of BMW cars, one using the system's CPU and one the GPU. Lower times are better.

The second test, SPECviewperf 2020, is a widely recognized testing suite that measures graphics performance for professional applications. The software renders, rotates, and zooms in and out of solid and wireframe models using viewsets from popular independent software vendor (ISV) applications. We run the 1080p resolution tests based on PTC's Creo CAD platform; Autodesk's Maya modeling and simulation software for film, TV, and games; and Dassault Systemes' SolidWorks 3D rendering package. Results are delivered in frames per second (fps), and higher numbers are better.

The Result? Authoritative Guides to Buying the Best Laptops

All this testing is a key part of how PC Labs informs the content and choices within our roundups and reviews on PCMag.com. Performance is just one factor in the overall picture we take of a laptop's worth, which also includes assessments of its design, feature set, value for money, build quality, and ergonomics.

We maintain more than two dozen guides to buying laptops in a variety of niches, large and small. Our main best laptops guide is the place to start for a wide-ranging overview of the top choices on the market across categories (as well as a deep explainer on the key facets of laptop buying). Other large laptop niches for which we update guides constantly include gaming laptops, business laptops, 2-in-1 laptops, and Chromebooks. We update these guides in sync with major releases to the market and, especially, when we bestow our Editors' Choice award onto a given machine.