This page compares the results of sustained cpu endurance tests on the selected VPS plans.
Endurance tests measure the capacity of a VPS to sustain intense and prolonged CPU utilization. Each test consist of a CPU intensive task run in a loop continuously for 24 hours. The tests targets an overall 50% CPU usage on the VPS.
|Provider||Plan||Cost||Number of CPU cores||Trial|
|Google Compute Engine||g1-small||$27.0||1||Feb 11 2018||Endurance Run details|
|Amazon EC2||t2-medium||$37.0||2||Jan 14 2018||Endurance Run details|
|Vpsdime||Premium 4GB||$20.0||1||Dec 18 2017||Endurance Run details|
|OVH||VPS SSD 1||$4.47||1||Dec 03 2017||Endurance Run details|
|Vultr||VC2 1GB $5||$5.0||1||Nov 27 2017||Endurance Run details|
This is the average number of iterations the VPS executed every hour over 24 hours.
A higher number means the VPS is faster.
The coefficient of variation represents the stability of the VPS performance over the course of the test. It's the standard deviation divided by the mean number of iterations per hours.
A lower number indicates better stability.
This is the average proportion of CPU that is "stolen" by the hypervisor. A lower overall performance can often be explained by cpu steal running over 1%.
A higher number means the VPS host may be overallocated.
This is the average proportion of CPU left idle during the test. It simply serves to verify that the target 50% cpu usage was reached.
hostbenchmarker.com: shared hosting benchmarks, where you can see results of speed tests, uptime reliability, and tech support response times.
All vpsbenchmarks.com trials are conducted without the knowledge and without sponsorship from the provider of the VPS where the benchmarked application is running.
To support VpsBenchmarks, make your hosting purchases using the Provider Affiliate Links displayed on the site.
To this days all first tier cloud providers have taken action against the Spectre and Meltdown ...
All tests that were conducted at VpsBenchmarks: