Download speed versus download time

The classic mistake is assuming that a 500 Mbps line means downloads happen at 500 megabytes per second. That is simply wrong. Bits and bytes are not the same thing, and real transfers also lose performance to overhead, Wi-Fi quality, storage speed, and server limits.

The core formula

Download time comes down to one relationship: divide the file size by the transfer speed and you get time. Written out:

Download time (seconds) = File size (MB) / Transfer speed (MB/s)

The complication is that file sizes and connection speeds are typically expressed in different units. Operating systems report file sizes in megabytes (MB) or gigabytes (GB). Internet providers advertise speeds in megabits per second (Mbps). These look similar but differ by a factor of 8, because one byte contains 8 bits.

To convert a connection speed from Mbps to MB/s, divide by 8. A 100 Mbps connection delivers 100 / 8 = 12.5 MB per second at the theoretical maximum. That is the number to use in the time formula, not 100.

A worked example with real numbers

Suppose you have a 100 Mbps broadband connection and you want to download a 5 GB game update. Here is the calculation step by step.

First, convert the file size to megabytes: 5 GB = 5 x 1024 MB = 5,120 MB (using binary gigabytes, as most operating systems report them).

Second, convert the connection speed to MB/s: 100 Mbps / 8 = 12.5 MB/s.

Third, divide: 5,120 MB / 12.5 MB/s = 409.6 seconds, which is about 6.8 minutes.

That is the theoretical best case with no overhead, no other traffic, and a server that can actually push at 12.5 MB/s. Real downloads are typically slower, which is covered below.

The Mbps to MB/s converter handles the first step, and the download time calculator handles all three steps together.

Why real downloads fall short of the theoretical figure

Several factors reduce actual throughput compared to what the formula predicts from your plan speed.

Protocol overhead. TCP/IP adds headers and acknowledgment packets to every data transfer. At typical Ethernet frame sizes the overhead is roughly 2 to 5 percent. It is small but constant.

Shared bandwidth. Home broadband is almost always a shared medium at some point between your router and the ISP backbone. During peak evening hours your available bandwidth can drop significantly below the plan rate.

Server-side limits. A download server may cap individual connections, prioritize paying subscribers, or simply be saturated. A 1 Gbps home connection cannot pull faster than a server that throttles at 50 MB/s.

Wi-Fi versus wired. A Wi-Fi connection on a 2.4 GHz band in a busy apartment building may deliver 20 to 40 Mbps even if your router is connected to a 500 Mbps line. Interference, distance, and competing devices all cut into the radio link. Wired Ethernet removes that variable.

Disk write speed. Very fast connections combined with slow storage can create a bottleneck at the drive. A spinning hard disk writing at 80 to 120 MB/s will cap your effective download rate if the connection can push more than that. A modern SSD typically writes fast enough that this is rarely the limiting factor for most consumer plans.

In practice, a reasonable expectation for a reliable wired connection is 70 to 90 percent of the advertised Mbps figure, converted to MB/s. On Wi-Fi, 40 to 60 percent is a more realistic working estimate.

Quick reference: approximate download times at 100 Mbps

These figures assume 100 Mbps (12.5 MB/s) and no overhead reduction, giving the theoretical minimum time. Add 20 to 50 percent for a realistic estimate on a typical home connection.

• 50 MB (a large PDF or photo album): about 4 seconds
• 1 GB (a short film or OS update): about 82 seconds (1.4 minutes)
• 5 GB (a console game or Blu-ray rip): about 410 seconds (6.8 minutes)
• 25 GB (a large PC game or 4K film): about 34 minutes
• 100 GB (a full game library backup): about 2.3 hours

On a 1 Gbps (gigabit) connection, divide all of those times by 10. On a 25 Mbps connection, multiply by 4.

Practical takeaways

Always divide your Mbps figure by 8 before estimating download time. Skipping that step inflates your speed estimate by 8x and makes every download seem faster than it is.

Use theoretical time as a floor, not a target. Plan for a transfer to take 1.3 to 1.5 times the formula result on a good wired connection, and up to twice as long on a congested Wi-Fi network.

For large transfers such as server backups, NAS copies, or game installs, run a short speed test with a tool that reports in MB/s rather than Mbps so you can plug that number directly into the formula without converting.

Server limits are outside your control. If a download consistently peaks well below your connection speed, the bottleneck is on the remote end, not your line.