Simply put, our performance cannot be beat. Our competition isn’t other VPN providers or protocols, it’s the line rate. No one can say that! Our superior network design really does show itself off with these results. We’ve looked at all the bottlenecks, and made sure we weren’t the problem. Even our containerization proved itself out. All tests were carried out using an independent third-party, speedtest.net.
Wired Connection
Windows 11 Professional running WireGuard NT kernel version and Meshify Agent. The NIC is a Realtek Gbe adapter connected to an Arris NVG578HLX router. The machine itself is a three year old I5-10400. Its only real requirement was the ability to saturate a 1Gbps connection, which it can. All equipment is consumer grade. No tuning was done before the test.
Interpreting the Results
In this case, there is no bottleneck. Anything above 80-90% of line rate is generally accepted as maxing out a connection. For most customers, this is where their ISP will be the bottleneck. Even highly rated internet gateways do not advertise more than 500Mbps WireGuard performance (which means they likely have a CPU bottleneck).
Wireless
For wireless we used a Windows 11 Professional Dell Inspiron 15 with an AMD Ryzen 5700U CPU and a Qualcomm Wireless 802.11ac adapter to the same Arris router. No turning was done prior to the test.
At this point, we’re simply measuring the performance of the wireless router (or the wireless card), as that is where the bottleneck is. I suspect if we got a little closer we could squeak out more performance. And that’s the point: you can actually use our service, and have no perceptible loss in performance. If you’re a regular user of OpenVPN, perhaps for work, you know that’s not normally the case with a VPN.
Latency is the predominate cause of lower performance. As you can see from the two ping times above, going from 12ms latency to 20ms latency cut performance in half. The mere milliseconds it takes to talk to your wireless router can dramatically impact high-end performance, where every millisecond counts.
Conclusion
We are a game changer for virtual private networks. Not only do we provide safe, anonymous browsing, but we give it to you with no loss in performance. Don’t be fooled by competitors with their hacked up go implementations of WireGuard. They can’t manage half our performance. And OpenVPN, well, we consider that our future market share.
Performance can vary widely based on equipment used. For example, using a Raspberry PI we could not push more than 200Mbps, with or without WireGuard, using a wired connection. That is absurdly bad in comparison, and we will need to investigate further before publishing those results.
Do you have some performance results to share? Send your screenshots to Meshify Support.