Benchmarks refer to the process of running a series of tests using a set of computer programs or operations in order to assess the relative performance of hardware and/or software. Benchmark is also related to the term benchmarking, which is the comparison of performance of an object to another to set a standard among various systems.




Benchmarks serve as an alternative to real applications if such applications are unavailable or if the administration cost are too high. There are two types of benchmarks system namely synthetic and application.  Synthetic benchmarks specifically provide numbers and data by comparing one system against other systems that are already benchmarked. The advantage of synthetic benchmark is that it can be replicated on several different systems. The disadvantage is that it does not go beyond the numbers and data when it comes to assessing the relative performance of an object.


Examples of Synthetic Benchmarks:


1.      Sisoft Sandra – A set of benchmarks combined together to figure out component performances and establish an overall system


2.      CineBench R10 – Used to check your computer that measures the performance to the main processor and graphics card


Application benchmarks are often called “real-world” programs because it provides information that is more practical to users. The advantage of using application benchmarks is that it is more accurate compared to synthetic benchmarks in terms of identifying the real performance level of hardware and/or software.  On the other hand, application benchmarks are much more difficult to interpret because it provides more subjective results. In addition, people can also have difficulty in understanding some of the technical aspects of the information.


Examples of Application Benchmarks:


1.      Encoding LAME – It is a software application that is used to change audio into an MP3 format


2.      DIVX Encoding- Used to watch HD and high-quality digital video in your computer.

[i]    Wikipedia Website.

ii   Techwarelab Website.