The position of a specific VPS tariff is determined by the VPS index, which is calculated based on the tests listed. The position of the hosting provider depends on the average index of all analyzed VPS.
All virtual machine performance tests are repeated three times. The test execution time is selected randomly within a range of 3 days.
This means that the hosting provider cannot influence the VPS test results in any way.
Performs performance analysis of one core and all declared cores on the VPS. Testing is performed using the Geekbench 6 benchmark.
As a rule, the performance of all cores is not equal to the performance of one core multiplied by the number of cores. For example, for a 4-core VPS, the indicators may be as follows: Multi-Core Score - 3585, а Single-Core Score - 1129.
1129 x 4 = 4516, which is greater than 3585
This means that performance decreases when all cores are loaded at the same time. Therefore, we take into account the percentage decrease in performance when creating the VPS index.
The Sysbench benchmark utility is used to analyze the speed of operations in random access memory (RAM). This is a multi-threaded open-source tool that analyzes the number of events performed per unit of time.
Disk system performance is analyzed in two stages:
The Fio utility is used with parameters selected to be as optimal as possible to eliminate the possibility of data caching and to approximate the actual load on the server as closely as possible.
The first stage determines the maximum number of operations per second. Two operations are performed to obtain IOPs for random read and random write. This stage is the most important because it generates the load that is most often seen on a VPS.
In the second stage we obtain the maximum possible data transfer speed during linear (sequential) reading and writing.
Network speed, as well as latency (response time), is checked using Speedtest. Testing is performed at several locations in Europe, Asia, and North America.
The VPS index takes into account not only the speed at a given moment, but also compares its change over time and compares speeds in different directions. The ideal network quality will be provided by a virtual server with approximately the same speed in all directions around the world and no significant changes in speed and latency.
It is also important to note that IndexVPS determines the VPS location and gives preference to locations closer to the server. However, it is tested in all directions.
IndexVPS's point of view is based on extensive experience working with servers, including virtual ones, as well as experience in programming and analyzing performance issues on various websites.
We believe that the most important factor is processor (CPU) performance, as no operation can be performed without passing through the processor's computing resources.
The second most important factor is disk system performance. Despite the fact that all operations are performed using RAM, just like with the processor, the importance of RAM performance cannot be overestimated. The speed of processing operations in memory naturally varies between servers, but RAM remains faster than disk storage.
Network quality is also important when determining the VPS index, since access to the virtual server is always remote. VPS is rarely used for performing exclusively local operations.
IndexVPS shows only reliable VPS performance ratings. Our VPS index cannot be purchased.
Advertising is sold exclusively in the advertising block and does not affect the index itself.