CCalclabhub

Technology

How To Calculate Download Time

Download time is file size divided by connection speed, provided both are expressed in the same unit. The trap is that file sizes are quoted in bytes while connection speeds are quoted in bits.

Quick Answer

time = size in megabits / speed in megabits per second

size
File size, converted to megabits by multiplying gigabytes by 8000
speed
Connection speed in megabits per second
t
Transfer time in seconds, then converted to minutes
eff
Fraction of the advertised speed actually achieved

A 20 gigabyte file is 160000 megabits, so a 100 megabit per second line takes 1600 seconds at best. Allowing for 92 percent real-world efficiency stretches that to about 1739.1304347826 seconds, which is 28.9855072464 minutes. The same connection delivers 12.5 megabytes per second.

What Is Download Time?

Download time is the file size divided by the transfer speed, and the whole difficulty of the calculation lies in getting the two quantities into matching units. File sizes are conventionally quoted in bytes while network speeds are conventionally quoted in bits, so a step of multiplication by eight sits between them whether you notice it or not.

One byte is eight bits. That single fact is the source of most confusion about download speeds, because an eight megabit per second line delivers exactly one megabyte per second, not eight. Dividing the advertised megabit figure by eight gives the throughput in megabytes per second, which is the number a file manager will show you.

Converting the file size into megabits is the other half of the job. A gigabyte is a thousand megabytes and each megabyte is eight megabits, so one gigabyte is eight thousand megabits. A 20 gigabyte file therefore weighs 160000 megabits, and on a 100 megabit per second line that is 1600 seconds of pure transfer.

That 1600 second figure is a theoretical minimum, and no real connection reaches it. Protocol overhead, packet acknowledgements, other devices on the network and the speed of the server at the far end all take a bite. Assuming a real-world efficiency in the region of 90 to 95 percent of the advertised rate gives a more useful estimate than the headline number.

The efficiency factor is applied as a divisor on the speed, or equivalently as a multiplier on the time. At 92 percent efficiency the effective rate is 92 megabits per second rather than 100, and the same 160000 megabits now takes 1739.1304347826 seconds. That is a little under half a minute per gigabyte of extra time compared with the ideal.

Dividing the seconds by sixty gives the answer in minutes, which is usually the more useful form. The 1739.1304347826 seconds above becomes 28.9855072464 minutes, comfortably under half an hour. Quoting the result in minutes also makes it easier to judge whether a download is worth starting now or scheduling for later.

The same arithmetic answers the reverse question: how much data a given time budget buys. Multiplying the effective speed by the time in seconds and dividing by eight thousand gives the gigabytes transferred. This is the calculation behind a data cap, and it is worth doing before leaving a large transfer running overnight.

Storage and memory use the same prefixes with a subtle difference. A gigabyte in the decimal sense is one thousand million bytes, while a gibibyte is two to the thirtieth bytes, which is about 1074 million. Operating systems often display the binary version while labelling it gigabytes, so the file size you read may already be a few percent larger than the decimal figure the arithmetic expects.

Connection speeds are almost never symmetric. Domestic broadband commonly offers a much faster download rate than upload rate, so the formula for sending a file uses a different speed from the one for receiving it. Using the download figure for an upload overstates the throughput, sometimes by a factor of ten.

Once the numbers are in matching units the arithmetic is a single division, and the only judgement left is choosing a realistic efficiency. A wired connection on a quiet network might reach 95 percent of the advertised rate; a congested wireless link might manage 60 percent. Picking the efficiency honestly is what turns a theoretical figure into a usable estimate.

Formula

t = size / speed

The core relation. Size must be in megabits and speed in megabits per second for the result to come out in seconds.

SymbolMeaning
sizeFile size
speedConnection speed
tTransfer time

megabytes per second = megabits per second / 8

Converts the advertised network speed into the unit a file manager displays. One byte is eight bits, so the megabyte figure is always the smaller number.

SymbolMeaning
MbpsSpeed in megabits per second
MBpsSpeed in megabytes per second

t_real = size / (speed x eff)

Divides the effective speed rather than the advertised one. An efficiency of 0.92 models overhead, contention and server limits in a single number.

SymbolMeaning
sizeFile size
speedAdvertised speed
effEfficiency

How To Calculate Download Time

  1. 1

    Convert the file size to megabits

    Multiply the gigabytes by 8000. A 20 gigabyte file becomes 160000 megabits, because a gigabyte is a thousand megabytes and each megabyte is eight megabits.

  2. 2

    Keep the connection speed in megabits per second

    The advertised figure is already in the right unit, so no conversion is needed at this stage. Converting it to megabytes now would double-count the factor of eight.

  3. 3

    Divide to get the theoretical seconds

    160000 megabits on a 100 megabit line is 1600 seconds. This is a floor, not a prediction, because it assumes every bit arrives at the full advertised rate.

  4. 4

    Multiply the speed by the efficiency

    Ninety-two percent of 100 megabits per second is 92 megabits per second. Using this effective rate in place of the advertised one gives the realistic transfer time.

  5. 5

    Convert the seconds to minutes and check the size

    Divide by sixty to get 28.9855072464 minutes. Cross-check by noting that the connection moves 12.5 megabytes per second, which at that rate delivers the file in roughly half an hour.

Examples

Example 1: A 20 GB file on a 100 Mbps line

File size
20 gigabytes
Speed
100 megabits per second
Efficiency
92 percent
StepCalculationResult
Size in megabits20 x 8000160000
Ideal seconds160000 / 1001600
Effective throughput100 x 0.9292
Realistic seconds160000 / 921739.1304347826
Realistic minutes1739.1304347826 / 6028.9855072464

Result: The transfer takes 1600 seconds at the advertised rate and about 1739.1304347826 seconds in practice, which is 28.9855072464 minutes.

Example 2: A 4 GB file on a 50 Mbps line

File size
4 gigabytes
Speed
50 megabits per second
Efficiency
100 percent
StepCalculationResult
Size in megabits4 x 800032000
Seconds32000 / 50640
Minutes640 / 6010.6666667

Result: Four gigabytes at a steady 50 megabits per second takes 640 seconds, which is 10.6666667 minutes.

Example 3: A 700 MB file on a 20 Mbps line

File size
0.7 gigabytes
Speed
20 megabits per second
StepCalculationResult
Size in megabits0.7 x 80005600
Seconds5600 / 20280
Minutes280 / 604.6666667

Result: A 700 megabyte file over a 20 megabit connection needs 280 seconds, which is 4.6666667 minutes.

Calculator

Realistic download time in minutes

28.9855

Ideal time in seconds
1,600
Realistic time in seconds
1,739.1304
Throughput in megabytes per second
12.5

Values update as you type. This calculator covers the single scenario its formula assumes — see Common Mistakes for what it leaves out.

Prefer a full-width tool? Open the Download Time calculator page.

Common Mistakes

  • Treating megabits per second as megabytes per second

    One byte is eight bits, so an eight megabit line moves one megabyte per second. Reading the speed as megabytes understates the download time by a factor of eight.

  • Forgetting to convert the file size into megabits

    Multiplying gigabytes by 8000 is the step that makes the units match. Skipping it leaves the size in bytes and the time comes out eight thousand times too large.

  • Assuming the advertised speed is what you get

    Headline rates are theoretical maxima. Overhead, contention and server limits typically remove five to forty percent, so an estimate without an efficiency allowance is optimistic.

  • Using the download speed for an upload

    Domestic connections are usually asymmetric, with upload far slower than download. Substituting the download figure for an upload can overstate the throughput by an order of magnitude.

  • Mixing up decimal gigabytes and binary gibibytes

    A gibibyte is about 1074 million bytes while a decimal gigabyte is exactly one thousand million. Operating systems often display the binary value under a gigabyte label, so the size you read may already be larger than the arithmetic assumes.

  • Dividing by sixty before the efficiency step

    The efficiency multiplies the speed, so it has to be applied while the time is still in seconds. Applying it after converting to minutes gives a result that is off by a factor of sixty.

FAQ

Why is my download slower than the speed I pay for?

The advertised speed is a theoretical maximum for the link itself. Protocol overhead, other devices sharing the connection, wireless interference and the upload capacity of the server all reduce the achieved rate, which is why an efficiency allowance of ninety to ninety-five percent is realistic on a good wired link.

How many megabytes per second is a 100 megabit connection?

Twelve and a half megabytes per second, because one byte is eight bits and 100 divided by 8 is 12.5. Download managers show that figure, which is why a plan described as 100 looks like only 12.5 in the software.

How long does a 1 GB file take to download?

At 100 megabits per second it takes 80 seconds at the theoretical maximum, because one gigabyte is 8000 megabits. Allowing for overhead pushes that to roughly 87 seconds in practice.

Does the calculation work for uploads as well?

The arithmetic is identical but the speed is different. Use the upload rate rather than the download rate, since domestic connections commonly offer far less upload bandwidth and substituting the wrong figure produces a badly optimistic answer.

Why do file sizes and speeds use different units?

Network engineers have always measured throughput in bits per second because the physical link carries a stream of bits, while storage has been measured in bytes because memory is addressed in bytes. The mismatch is historical, and it is the reason the factor of eight has to be handled deliberately.

What efficiency figure should I use?

About 95 percent on a quiet wired connection, 85 to 90 percent on a good wireless link, and lower still when the network is busy or the server is slow. Choosing the figure honestly matters more than the precise value, because the ideal figure is never achieved.

References

  1. [1]Wikipedia, Data-rate units — https://en.wikipedia.org/wiki/Data-rate_units
  2. [2]Wikipedia, Bit rate — https://en.wikipedia.org/wiki/Bit_rate
  3. [3]Wikipedia, Bandwidth (computing) — https://en.wikipedia.org/wiki/Bandwidth_(computing)