Bandwidth Calculator — Transfer Times
Bits and bytes have been confusing each other’s customers for forty years: connections are sold in megabits, files are measured in megabytes, and the factor of eight between them is where download estimates go to die. Enter a file size and a connection speed, and this calculator returns the transfer time with the conversion done honestly and the conventions stated — plus the data a stream consumes per hour at that speed.
The Factor of Eight, and Its Close Cousins
Every bandwidth error in the world is one of three confusions, and this calculator is built to refuse all of them. Bits versus bytes: speeds are quoted in bits (Mbps), sizes in bytes (MB, GB), and a byte is eight bits — so a 100 Mbps line downloads at a theoretical 12.5 MB per second, the single most re-learned fact in consumer technology. Decimal versus binary: the industry sells gigabits in powers of ten while operating systems count gibibytes in powers of two (1 GiB = 1,024 MiB ≈ 7.4% more than 1 GB across a full unit ladder), so a “100 GB” game on a “1 Gbps” line resolves differently depending on whose units hold the chalk — the calculator states its convention in its note, which is more than most marketing does. Headline versus sustained: the rated speed is a peak under laboratory manners; the note’s 70–90% realism band is the honest planning figure, and the gap between the calculator’s floor and reality is a feature — it measures your connection’s manners, not your arithmetic’s error.
Worked Example: A DVD’s Worth at Broadband Speed
The defaults price a familiar errand: 4.7 GB (a single-layer DVD’s capacity, and a fair stand-in for a chunky game update) over a 100 Mbps connection. Bits to move: 4.7 × 8,000 = 37,600 megabits. Time = 37,600 ÷ 100 = 376 seconds — 6 minutes 16 seconds at the full sustained rate, or about seven to nine minutes in the lived world of the note’s realism band. The second line scales the same pipe: an hour at 100 Mbps moves 45 GB — the figure data caps are denominated against (a 1.2 TB monthly cap is about 27 hours of full-throttle line, a comparison that reorganises how a household thinks about “unlimited-feeling” broadband). One division, honestly converted, prices both the errand and the month.
Upload: The Forgotten Direction
The same formula with the asymmetry of real connections is most users’ actual bottleneck: consumer broadband is sold by its download figure while uploads run at a tenth or twentieth of it, so the 4.7 GB that downloads in six minutes can take an hour or two to send — the arithmetic behind every creator’s overnight upload and every cloud backup’s first-weekend saga. Run this calculator with your upload speed before scheduling a delivery, and run it again with the client’s or server’s cap if theirs is lower; a transfer moves at the slowest of the sender’s up, the receiver’s down, and every policy limit between. The formula never changes — only which number deserves to be in the speed box, which is why the box is yours to fill rather than a broadband advertisement’s promise to fill on your behalf.
Planning a Network Purchase With This Arithmetic
The calculator’s division runs a household’s and a small office’s upgrade decisions in both directions. Forward: given the line, what do the month’s real loads cost in time — the designer’s 40 GB library sync, the household’s game downloads, the backup window’s nightly increment — and do they fit the hours available? Backward: given the load and the deadline, what speed must be bought — a studio that must move 2 TB nightly inside a 6-hour window needs 2,000 × 8,000 ÷ 21,600 ≈ 740 Mbps of sustained throughput, which prices the business tier and the symmetric upload product in one line, and explains why such studios colocate storage instead. The ISP’s headline tiers are marketed in the forward direction only; run the backward division before signing, and the contract will be the first one that fits the workload rather than the advertisement.
Latency’s Shadow: Why Small Transfers Ignore This Page
The division rules bulk, but the web’s everyday traffic is not bulk: a page load is hundreds of small requests, each paying latency tolls (DNS, connection setup, TLS, round trips) that dwarf their transmission time, so a gigabit line and a 100-megabit line can render the same site near-identically while differing a hundredfold on this calculator’s errands. The planning corollary cuts both ways — do not buy bandwidth to fix a latency complaint, and do not blame latency for a download estimate: know which regime your problem lives in. The calculator’s silence about milliseconds is the honest boundary of its competence, marked here so the reader can route each frustration to its rightful remedy.
A Note on Megabits in the Wild
The Mbps figure itself deserves one sceptical paragraph: it is measured under the provider’s chosen conditions, against the provider’s chosen servers, in the direction the provider prefers to advertise. International routes, peering disputes and a congested last mile all tax it differently, which is why the same line that tests at 940 Mbps domestically delivers a transatlantic backup at a tenth of that. The calculator cannot know your route; it can only insist that the speed entered be the speed measured on the path the transfer will take, at the hour it will take it. Given that honesty at the input, its one division has never yet been the part of the estimate that failed.
Related Tools
More from the Other Calculators hub: the Upload Time Calculator — this page’s sibling dedicated to the sending direction — and the Uptime Calculator for the availability of the service at the other end.
Frequently Asked Questions
Why is my download slower than this calculator says?
The calculator prices the pipe at its rated, sustained best; reality subtracts its tolls in order: protocol overhead (TCP/IP and encryption take their few percent), the server’s own rate limits and congestion (the commonest cap on big-platform downloads), Wi-Fi contention and distance, and your provider’s peak-hour shaping. Measure with a speed test at the same hour you plan to transfer, enter that figure rather than the package’s headline, and the calculator’s floor becomes a forecast. A persistent gap between wired and Wi-Fi results is a home-network diagnosis waiting to happen, not an ISP conspiracy.
How much data does streaming video use per hour?
Read it off the bitrate with the calculator’s second line: data per hour = bitrate × 0.45 GB per Mbps. Standard-definition streams run about 1–3 Mbps (0.5–1.4 GB/h), HD about 5–8 Mbps (2.3–3.6 GB/h), and 4K about 15–25 Mbps (6.8–11.3 GB/h), varying by platform and scene complexity. Against a monthly cap the multiplication is the whole budget: a household streaming 4K for four hours a night spends roughly 1 TB a month on that habit alone — the worked example’s framing, scaled to the sofa.
What is the difference between bandwidth and speed?
Strictly, bandwidth is the pipe’s capacity (the maximum rate, what the calculator consumes), “speed” is the experienced rate of a particular transfer, and latency — the delay before data starts flowing — is a third quantity entirely, invisible to this arithmetic and decisive for gaming and calls. A high-bandwidth, high-latency link (satellite) downloads films beautifully and plays shooters terribly. This page prices bulk transfer, where capacity rules; small interactive loads live and die by the milliseconds no file-size division can see.
How long to download a 100 GB game?
100 × 8,000 ÷ your Mbps, in seconds: at 100 Mbps, 8,000 seconds — about 2 hours 13 minutes at the impossible best, nearer 2.5–3 hours lived; at 300 Mbps, about 45 minutes ideal; at gigabit, about 13 minutes ideal, whereupon the bottleneck usually emigrates to the console’s storage write speed and the platform’s servers. The formula scales linearly and mercilessly, which is why preload culture exists: the only way to beat the division is to start it earlier, and every platform’s pre-download button is an admission of the arithmetic on this page.
Do VPNs change the calculation?
They change the sustained rate, not the arithmetic: encryption overhead and the detour through the VPN’s servers typically cost 5–20% of throughput (more on distant endpoints), so enter your speed as measured through the VPN and the calculator’s answer holds. The same principle covers every middlebox — proxies, corporate gateways, a household of simultaneous streamers dividing the pipe: the speed in the box must be the share your transfer will actually get, and halving the headline for a busy house is realism, not pessimism.
Why do speed tests show different units and numbers?
Tests report in Mbps by convention but some display MB/s (the eight-fold trap again, in reverse: a test “only” showing 12.5 is your healthy 100 Mbps line in byte dress), and results vary by server, protocol (single versus multi-connection) and minute. For this calculator’s purposes, run two or three tests, take the typical sustained figure in Mbps, and prefer a wired measurement when the transfer will be wired, since a wireless measurement of a wired job answers a question nobody asked. Precision beyond that belongs to network engineering; the planning question — minutes or hours? — resolves at far coarser resolution, and it is the question this page was built to answer.