How We Test and Rank Laptops
It is crucial to always carry out standardized, critical, and consistent testing on laptops to ensure that your readers are well informed about the latest machines in the market.
Also, thorough research about a laptop means that readers and anyone interested in getting a laptop will be able to trust your reviews.
Our laptop testing concept is in three major categories: productivity testing, graphics testing, and a battery-life trial.
These three categories will dictate the overall performance level of a laptop. After we have carried out these three tests, we can now give you a final verdict of how good a laptop is.
Below, I will explain these three categories.
This is, of course, one of the most important aspects, where we determine the machine’s everyday productivity level by using the widely accepted UL’s PCMark 10 benchmark program, which has been developed to simulate how we use our laptops daily.
PCMark 10 will evaluate the laptop’s overall performance level when used for general computing tasks like word processing, surfing the internet, spreadsheet tasks, video conferencing, and some other ones. The test will come up with a numeric score: the higher the score means the better the overall performance of the laptop for general computing tasks.
Let me also point out that a higher screen resolution will have a major impact on the laptop’s performance on PCMark 10.
The next step is to determine the speed of the system’s main hard drive using another UL program which is the PCMark. Similar to PCMark 10 that we used to assess the performance level, the PC Mark 8 also uses the numeric score to determine the booting speed of the hard drive, and the higher numbers are always better.
It’s common knowledge that laptops with solid state drives (SSDs) are always better faster than HDDs. So you can expect a laptop equipped with an SSD to score higher than a laptop with an HDD.
The next test is to assess the processors’ capability by using Maxon’s CPU-crunching Cinebench R15 test. To ensure that we have a perfect result, we carry out this test at the All Cores setting. This setting is a complete processor horsepower test.
It means that the test will make use of all the available processors in the system to get the correct result.
Cinebench focuses on the processors and not the graphics card to deliver a reliable score. It will determine whether the laptop is powerful enough to handle complex power-intensive tasks when used with a high-end program.
Adobe Photoshop CC Photo Editing Test
This is the final productivity test that we always run on laptops, and it is done by using a modified Adobe Photoshop image-editing program. We make use of a bit old Creative Cloud version of Adobe Photoshop to run different tasks on a basic JPEG image.
For example, we will pick a picture and apply several resource-intensive features of the program, such as Watercolor, Mosaic Tiles, Stained Glass, and blur effects. These effects and filters consume a lot of the system’s resources, which makes the program to be considered a valuable tool for the test.
The test stresses every major component of your system: CPU, storage subsystem, memory, and even the GPU.
Evaluating the graphics performance of a system requires using different challenging tests to produce different meaningful results. You can now put these different results together to determine the overall graphic performance level of a system.
For example, some tests will deliver results in scores with the higher score being the better, while there are some that you will use to determine the frames per second at which the GPU produces frame in a sequence, which means the quality of the image on the screen.
Synthetic Tests: 3DMark and Superposition
The first test to determine the synthetic graphic performance of a system is UL’s 3DMark. It comprises several subsets that evaluate the level of graphics of a system by rendering sequences of heavy 3D graphics.
There are two different types of 3DMark subtests, Sky Diver (perfect for laptops and budget PCs) and Fire Strike (capable of handling workstations). The result is also measured in numeric scores.
The second synthetic graphic test is the Superposition test: similar to 3DMark, this test renders and scans through a detailed 3D scene to evaluate the laptop’s performance. The only between these two tests is that Superposition works in the Unigine engine of the test, which offers a different 3D workload than 3DMark.
We carried out different Superposition tests, with resolution settings of 720p Low and 1080p High. In the end, we could deduce that for budget laptops, the reasonable target is around 30 fps, while high-end laptops can go as far as 60fps.
Real-World Gaming Tests
Both tests carried out above can help evaluate the general 3D graphics performance of a system. However, the most realistic way to know the actual graphic performance level of a system is to use modern computer games.
There are few latest laptop games like Far Cry 5 and Rise of the Tomb Raider that can be used to evaluate the graphic performance level of a system. These games can be used to assess how a laptop handles real-world games at different screen resolution settings.
We test these games on both medium and high graphics quality: we run the games on a resolution of 1080p if possible and see how the laptop copes. If the laptop can handle these games with a graphic s quality setting of 1080p without any color disruption, it means that the laptop belongs to the premium models.
This is the main reason why several top tech review sites will advise you to go for a high-end gaming laptop if you want to enjoy the latest AAA games.
Battery Life Testing
The final category is the battery life testing phase: this involves carrying out a video-playback-based battery rundown test that is compatible with all operating systems. This will give us an idea of how long the laptop will work without charging it.
The rundown test we always carry out on laptops is a looped video of Tears of Steel in MP4 format and 720p settings. If the laptop that we are testing has limited storage, we will play the MP4 file from an external device.
Before we begin the rundown test, we switch the laptop to a power-saving mode, and we reduce the screen brightness to 50 percent. We also make sure that the adaptive screen brightness feature of the laptop is deactivated.
W e also deactivate all the inbuilt innovative features that come with the laptop, such as wireless radios, keyboard backlighting, and any case lighting. We then start the video to see when the laptop will hibernate.
If the machine comes with two battery backup, we will carry out a separate test on each battery and then perform the test again on both batteries.
Evaluating battery performance can be quite challenging since results can vary widely depending on the pattern of usage. For example, a battery that is being used for general computing tasks will last longer than a battery for heavy gaming.
However, we believe that a laptop will be able to handle all-day general computing tasks if it can last for more than eight hours on our battery life test.
Special Cases: macOS, Chromebooks, Workstations
I should point out that we do not carry out all the above tests on every system. For example, we only run graphic-intensive games like Tomb Raider on high-end machines that can handle these types of programs.
Also, we do not use PCMark, 3DMark, or Superposition on Apple laptops because they are not compatible with macOS.
When it comes to Chromebooks, the only test that we can run is the battery rundown test, solely because it is the only one that is compatible with its operating system. To carry out this test on Chromebook, we use a lower resolution file, for example, a DVD rip of the complete Lord of the Rings trilogy, looped and saved on the Chromebook internal storage, if possible.
The next stage is to carry out the benchmarks CrXPRT and WebXPRT: this will make it easier for us to know the better Chromebooks when it comes to the battery backup. These benchmarks are very easy to use, and they report back in numeric scores to evaluate the overall battery life of a Chromebook.
All the tests that I’ve discussed above are compatible with workstations: we only need to modify the tests. These modified tests include the multimedia rendering tool POV-Ray that is used for virtual ray-tracing.
We also conduct other tests such as the SPECviewperf 13 suite that comes up with three “view sets” for the apps Creo, Maya, and SolidWorks, to see how a workstation will be able to handle the modification of relevant files in these three applications.