Work out how much bandwidth a site needs from users, application mix, concurrency and headroom, with a latency check for voice, video and VDI.
Buying too little bandwidth makes every video call and cloud app suffer, and buying too much wastes money every month. This calculator estimates the circuit a site needs from the number of users, what they run, how many are active at the busy hour, and how much headroom you want to keep spare.
Each application has a planning rate per active user: 0.1 Mbps for voice, 2.5 Mbps for video conferencing, 0.5 Mbps for SaaS and web, 5 Mbps for backup and file transfer, and 1.5 Mbps for virtual desktops. Enter how much of each a typical user generates, and the tool multiplies by users and the share active at once to get busy-hour demand, then divides by one minus your headroom to size the circuit. With the defaults of 100 users, 30% concurrency and 25% headroom, the result is about 67 Mbps.
If the round-trip time to your data center or cloud region is over 150 ms, the tool warns you that voice, video and virtual desktops are likely to feel slow regardless of bandwidth. The rates are planning assumptions. Measure your own traffic if you can, and size separately for peak events such as backups and software updates.
The calculator turns people and applications into a circuit size in three steps.
With the defaults of 100 users, 30% concurrency and 25% headroom, the demand per active user is about 1.67 Mbps, busy-hour demand is about 50 Mbps, and the circuit comes out at about 67 Mbps.
The planning rates per active user are 0.1 Mbps for voice, 2.5 Mbps for video conferencing, 0.5 Mbps for SaaS and web, 5 Mbps for backup and file transfer, and 1.5 Mbps for virtual desktops.
For each application, enter how many streams a typical active user has. A value of 1 means one stream per active user, and 0.2 means one in five active users has one. These are planning figures. Your own traffic may be lighter or heavier, so measure it where you can.
If the round-trip time to the data center or cloud region is over 150 ms and the mix includes voice, video or virtual desktops, the calculator warns that those applications may feel slow whatever the bandwidth. Also size separately for events that spike traffic, such as backups and software updates, check the upload direction on asymmetric circuits, and plan what happens when the main link fails.
It depends on what people do. With the default application mix, an active user needs about 1.7 Mbps. Heavy video conferencing, virtual desktops or large file transfers raise that, and light web and email use lowers it. Enter your own mix for a better estimate.
Concurrency is the share of your users who are active at the busiest time of day. If 100 people work at a site but only 30 are using the network heavily at once, concurrency is 0.3.
A link running close to full adds queueing delay and drops packets, which hurts voice and video first. Headroom leaves room for bursts and growth. The calculator divides demand by one minus the headroom you choose.
Yes. This estimate treats demand as a single figure, which is a reasonable start for symmetric circuits. On asymmetric connections, check that the upload speed can carry video calls, backups and cloud sync.
Not directly, but high latency makes voice, video and virtual desktops feel slow regardless of bandwidth. The calculator warns you when the round-trip time is over 150 ms and those applications are in the mix.
If the site cannot work without the network, plan a backup path sized for at least the critical traffic. The calculator sizes the primary circuit only, so run it again with just the essential applications to size the backup.
Per-call and total VoIP bandwidth for G.711, G.729, G.722 and Opus, including Ethernet, VPN and cRTP overhead, plus an estimated MOS from measured latency, jitter and loss.
Calculate network latency, ping distance, and data transfer times between geographic locations. Free tool for network engineers.
Estimate how long a file takes to download or upload, and find out whether the link, the TCP window or packet loss is the bottleneck.
Estimate round-trip latency between AWS, Azure and Google Cloud regions, or find the best region for your users.
Estimate voice call quality (MOS) from round-trip time, jitter and packet loss for G.711, G.729 and Opus.