ISP plans quote megabits per second. File managers show megabytes. The gap of eight, plus real-world utilization, is why “100 Mbps” rarely finishes a 100 MB file in one second. This guide peels ideal clock vs effective clock, the Mbps ↔ MB/s twin, and a speed ladder for one file. Open the Download Time Calculator for the live twin clocks.
What this guide owns
- Bits vs bytes (Mbps vs MB/s).
- Ideal time from size and rate; effective time after utilization.
- A speed ladder for the same file at different link rates.
- Handoffs to data-size and bandwidth-cost tools.
It is not a credit-score APR peel and not a unit-conversion habit desk.
The Mbps ↔ MB/s twin
1 byte = 8 bits MB/s ≈ Mbps ÷ 8 Mbps ≈ MB/s × 8
| Advertised link | Ideal payload rate |
|---|---|
| 25 Mbps | ~3.125 MB/s |
| 50 Mbps | ~6.25 MB/s |
| 100 Mbps | ~12.5 MB/s |
| 300 Mbps | ~37.5 MB/s |
| 1 Gbps | ~125 MB/s |
“Up to” on a plan is a ceiling, not a promise that every download saturates it.
Ideal clock vs effective clock
Ideal math (decimal sizes for a clean ISP-style sketch):
time (seconds) = (file_bytes × 8) ÷ (Mbps × 1,000,000)
Or with MB and MB/s:
time (seconds) = file_MB ÷ (Mbps ÷ 8)
Effective clock applies a utilization fraction because Wi-Fi contention, VPN overhead, TCP ramp-up, and busy origins rarely deliver 100% of the plan rate for the whole file:
effective time ≈ ideal time ÷ utilization
Example utilization bands (planning intuition, not physics):
| Condition | Rough utilization to try |
|---|---|
| Wired, quiet LAN, nearby CDN | 0.85–0.95 |
| Typical home Wi-Fi, evening | 0.60–0.80 |
| VPN + congested path | 0.40–0.70 |
On the calculator, read both clocks so you plan dinner around the effective number, not the brochure.
SI vs IEC file sizes (quick flag)
Many OSes label “GB” while counting GiB (2³⁰ bytes). Marketing and some download pages use decimal GB (10⁹ bytes). A “50 GB” label can disagree with itself by about 7%. For renaming peels, use the Data Size Converter. Below, examples use decimal GB/MB so the arithmetic stays transparent.
Worked example A: 50 GB game at 100 Mbps
File = 50 × 10⁹ bytes Bits = 50 × 10⁹ × 8 = 4 × 10¹¹ Ideal time = 4×10¹¹ ÷ (100 × 10⁶) = 4000 s ≈ 66.7 minutes
At 70% utilization:
Effective ≈ 4000 ÷ 0.70 ≈ 5714 s ≈ 95 minutes
Same file, speed ladder (ideal, decimal GB):
| Link | Ideal time |
|---|---|
| 25 Mbps | ~4.4 hours |
| 50 Mbps | ~2.2 hours |
| 100 Mbps | ~1.1 hours |
| 300 Mbps | ~22 minutes |
Worked example B: 500 MB deck at 50 Mbps
Ideal MB/s = 50 ÷ 8 = 6.25 Ideal time = 500 ÷ 6.25 = 80 s ≈ 1.3 minutes
At 60% Wi-Fi utilization ≈ 2.2 minutes. Still short, but the same derate logic scales to multi-hour game installs.
Mistakes that inflate disappointment
- Treating Mbps as MB/s (off by 8×).
- Using plan peak instead of a measured evening speed.
- Ignoring that the bottleneck may be the server or Wi-Fi, not the ISP modem.
- Starting a huge download on a metered cellular plan without a size check.
- Comparing SI “GB” marketing to IEC “GiB” OS progress without noticing.
How to run the calculator in five minutes
- Note file size (and whether the source meant GB or GiB).
- Enter a measured Mbps from a recent test at the same time of day, or your honest typical rate.
- Read ideal time, then set a utilization (or effective rate) for Wi-Fi/VPN reality.
- Climb the speed ladder if you are deciding whether an upgrade changes the evening.
- For bulk transfer cost questions, hand off to Bandwidth to Cost.
Open the tool: Download Time Calculator.
When this estimate is enough (and when it is not)
Enough for planning game installs, ISO images, and large work archives. Not enough as proof in an ISP dispute, SLA claim, or network-engineering capacity plan. Those need controlled tests, traceroute/path data, and contract language.
FCC consumer broadband context: broadband speed guide.
FAQ
Why is my download slower than a speed test?
Speed tests often hit optimized servers on short paths. Your file host may be farther, busier, or rate-limited. Wi-Fi and VPN add another ceiling.
Should I use MB/s or Mbps in the calculator?
Either works if the twin conversion is applied once. Do not mix them in the same formula without dividing or multiplying by eight.
Does parallel downloading always help?
Sometimes, until you saturate the path or trip server limits. Parallelism cannot beat a hard last-mile cap forever.
Is 1 Gbps eight times faster than 125 MB/s?
1 Gbps ideal payload rate is about 125 MB/s. They describe the same ceiling in different units.
Bottom line
Divide by eight to move from Mbps to MB/s, compute an ideal clock, then derate for real utilization. Run the same file on a speed ladder before you blame the progress bar. The Download Time Calculator keeps ideal and effective clocks side by side so planning matches the network you actually have.