Whenever you are buying a new laptop, a piece of PC hardware or a smartphone; you want to make sure you’re going to get the best performance for your money.
You can look at the specs, or get opinions from existing users, but for the most detailed insight you need to refer to the benchmark tests.
There are loads of technical sites on the Internet that deal in benchmarks. They’ll throw a whole host of charts and numbers at you for almost any piece of hardware on the market.
But what do they mean? And how can you use this information to make an informed decision?
Let’s take a look.
What Is Benchmarking?
Benchmarking involves running a series of software tests on a piece of hardware that replicate the kinds of tasks that it would perform in real world usage.
So, a CPU in a laptop will be subjected to assorted mathematical tests, and to measure how quickly it can compress or encrypt data. A hard drive would be tested on the speed it can write a single, very large file as well as thousands of very small files.
And benchmarks for the GPU (the graphics card) would measure things like the frame rate achieved while rendering different numbers of objects on screen at different levels of complexity and different resolutions.
The results don’t mean much on their own. But once you have subjected two products to the same tests then you can compare the results and begin to judge which device offers better performance.
How To Benchmark Your Own Gear
If you want to try it out for yourself to see how your own kit compares, there are plenty of apps available that can do the job.
The easiest app to get started with is Geekbench. This is a cross-platform tool that works on Mac, Windows, iOS and Android, and so enables you to compare performance across operating systems.
Geekbench focusses on processor and memory performance. For more detailed tests of other components like graphics and drive speeds, NovaBench and PassMark is good for desktop devices, and AnTuTu works well for Android phones.
Note that there is a huge number of variables that can affect the results of benchmark tests. Even something as simple as the level of charge in your laptop or smartphone battery could skew the results.
Most serious organisations that do benchmarking will take the effort to minimise these variables. Don’t be surprised if your kit — having been subjected to many months of real world use — produces different results.
Common Benchmark Tests And What They Mean
Benchmark tests differ based on the software used, or the person doing the testing. Since they will often focus on different areas, you can use multiple tests to get a bigger picture of how the hardware performs.
Here are some of the most common tests, what they mean, and what to look out for.
- Floating point math tests: Used by most tools, this tests the processor’s ability to perform a series of basic mathematical functions. A floating point indicates the numbers used involve fractions — integer tests using whole numbers are also tested separately. Results are often displayed in milliseconds, so lower numbers indicate faster performance.
- Compression tests: These test the speed at which the processor is able to compress large blocks of data in a lossless way. The results may be shown as a speed, in kilobytes per second, so a higher number is better.
- Single core tests: Used by services like CineBench or PassMark, these tests focus on the performance of a single core in the processor. Good single core performance is essential, as a lot of software is not optimised for multi-core processing.
- 2D graphics tests: 2D graphics tests focus on drawing, moving and scaling lines, fonts, elements within a user interface and so on. It’s often measured in frames per second, so a higher number is better.
- 3D graphics tests: A major test for gaming and graphics intensive applications, and used by tools such as Heaven Benchmark (whose test you can see in the video above) or 3DMark. These tests involve the rendering of few or very many 3D objects on screen, at varying levels of complexity, including detail, shadow, anti-aliasing and more, as well as testing different APIs (such as DirectX and OpenGL).
- Sequential tests: Hard drive benchmarks often focus on sequential read and write speeds, and random read and write speeds. Sequential refers to files being stored in a single chunk on the drive, such as a large file written to a non-fragmented hard drive. Results may be displayed in MB per second, so a higher number is better.
- Random tests: These test how the drive performs when required to access lots of data stored randomly throughout the drive. Random read and write times will be considerably slower than sequential times.
Smartphone-centric benchmark apps will test most of the same elements as those for desktops and laptops. But they also include a few extras, such as:
- SD card read/write speed: Similar to hard drive tests, this determines the speed at which data can be read or written to a memory card (or internal storage). As with hard drives, it’s measured in MB/s, so a higher number indicates faster performance.
- Database IO: Used in most smartphone benchmarking apps like AnTuTu, this measures the speed of reading and writing to the database on the device. Slow performance here can affect the overall performance of a device.
How Important Are Benchmarks?
Benchmarks are useful as a guide to hardware performance, but they are not the be all and end all. They are particularly good at showing how performance improves from one generation to the next, and can help you gauge value for money of a product, since you can easily see how it compares to similarly-priced alternatives.
Benchmarks are arguably best used when you have specific requirements, be it gaming, video editing or anything else that requires plenty of power to drive.
But for everyday computing tasks — surfing the web, Facebook, using Office — the performance differences are barely noticeable. Especially when there are other factors that can affect the speed of your computer.
Do you look consider benchmark results before you buy a phone or laptop? What other factors do you look at when choosing which product to buy? Let us know in the comments below.
Image credits: Main image via Mike Powell