Computers perform differently even if they have the same specifications. It depends on a lot of user factors such as hard drive fragmentation, installed software, viruses and a whole lot more. The performance of computer parts are usually measured not only by examining its features but how they compare with one another. This is what we call a benchmark. For example, in a graphics card benchmark, different video cards are tested using the same computer specs with the same applications and settings to see which performs better. A higher frame rate means that the video card is performing better. One of the reasons why these are tested would be for playing video games.
Amongst all PC applications and software, PC video games usually require the most power from your computer especially for newly-released state-of-the-art games. There are certain applications that can test your PC’s relative performance to other PC specifications. Some of them are specialized in such a way that they have specific tests for specific video card features and some perform benchmarks using the games themselves. They measure the frame-rate while running a demo of a particular game.
For example, Crysis Warhead is one of the most common benchmarks because of its rigid technology. A lot of powerful PCs can’t even make it playable at high resolutions when all of the advanced graphics features of the game are turned on. But what are the advantages and disadvantages of using video games as a PC performance benchmark?
One advantage of benchmarking PC performance with a particular game is if you really like that game. If you’re going to play your favorite game, of course, you would like it to run smoothly on your PC. As much as possible, you need it to be fast, seamless and of high quality. If you don’t have a suitable PC for a game that you like, you can reference benchmarks to see what hardware you’ll need not just based on paper (system requirements) but on actual performance. That would set your expectations more accurately.
However, if you’re not particular with the games, in-game benchmarks may yield inaccurate results. Games have different graphics engines meaning they are not programmed the same way. It usually depends on the genre of the game. This just simply means that if, for example, video card A runs faster than video card B on Game A, it doesn’t necessarily mean that video card A will run faster than video card B on Game B.
There are also quite a lot of in-game benchmarks that favor certain types or brands of video card. I have observed this in quite a number of game benchmarks. This is usually the result of a game using a technology specific to a video card brand’s feature in its graphics card lineup.
To make in-game benchmarks more effective, try to test the hardware on as many games as possible. You can combine old games with newer games that utilize new graphics technology. This way, you’ll see how your hardware fares with different game types, genres, engines and AI. This is very helpful in determining what upgrades you want to buy for your PC. Of course, as with any other computer user especially for game enthusiasts, you’ll want your computer to perform well regardless of what game you’re playing.
Wikipedia – Game Engine