Buy a “128GB” phone and Settings will tell you something closer to 113–119GB is actually free. Plug a “1TB” external drive into a laptop and Windows will cheerfully report 931GB. Pay for a “100 Mbps” broadband plan and your download manager crawls along at 12–13MB/s, not 100MB/s. None of these are faults, refunds, or scams — they’re two completely different number systems colliding, and almost nobody explains the actual math behind it.
This guide walks through exactly how digital storage and data-transfer units convert into one another, why manufacturers, operating systems, and internet providers each use a different base for their numbers, and three fully worked examples so you can check any device or plan of your own.
The Quick Conversion Table (Decimal Units)
Storage manufacturers, SIM/broadband providers, and most spec sheets use the decimal system — the same base-10 system as rupees and kilometres, where each unit is exactly 1,000 times the one before it.
| Unit | Symbol | Bytes (decimal) | Written out |
|---|---|---|---|
| Byte | B | 1 | 1 |
| Kilobyte | KB | 1,000 | One thousand |
| Megabyte | MB | 1,000,000 | One million |
| Gigabyte | GB | 1,000,000,000 | One billion (100 crore) |
| Terabyte | TB | 1,000,000,000,000 | One trillion (1 lakh crore) |
| Petabyte | PB | 1,000,000,000,000,000 | One quadrillion |
That “1 lakh crore” next to a terabyte isn’t a rounding trick — it’s exact. A terabyte is 1012 bytes, and 1 lakh crore is also 105 × 107 = 1012. If you’ve ever seen a government budget announced in lakh crore rupees, you already have an intuitive feel for how big a terabyte of data really is. (If large-number notation like 1012 looks unfamiliar, our guide to scientific notation on a calculator walks through reading and entering exponents like this, and our numbers-to-words guide covers converting between the Indian lakh/crore system and the international million/billion/trillion system in general.)
Why Your Device Disagrees With the Box: Binary vs Decimal
Computers don’t actually count in tens — they count in powers of two, because that’s how binary circuits work. So long before “kilobyte” meant exactly 1,000 bytes, early computer engineers borrowed the word “kilo” loosely to mean the nearest round power of two: 210 = 1,024. That habit stuck for decades inside operating systems, file managers, and RAM specs, even though hard drive and flash storage manufacturers (and the official SI/IEC standards bodies) always meant exactly 1,000.
To end the ambiguity, the International Electrotechnical Commission (IEC) created separate binary-prefix names in the 1990s — kibi, mebi, gibi, tebi — so “1,024 bytes” has its own correct label that never means 1,000.
| Decimal unit (manufacturers, SI) | Bytes | Binary unit (some OS, IEC name) | Bytes | Gap |
|---|---|---|---|---|
| 1 KB | 1,000 | 1 KiB (kibibyte) | 1,024 | 2.4% |
| 1 MB | 1,000,000 | 1 MiB (mebibyte) | 1,048,576 | 4.9% |
| 1 GB | 1,000,000,000 | 1 GiB (gibibyte) | 1,073,741,824 | 7.4% |
| 1 TB | 1,000,000,000,000 | 1 TiB (tebibyte) | 1,099,511,627,776 | 10.0% |
Notice the gap grows at every step — that’s the reason the storage “shortfall” feels worse the bigger the drive. A memory card barely feels the difference; a multi-terabyte NAS can look like it’s “missing” 10% of its advertised space purely from this unit mismatch, before a single file is even saved on it.
Most operating systems historically displayed binary values (1,024-based) but mislabeled them with the decimal symbol “GB” instead of the correct “GiB.” Apple’s Finder switched to genuinely decimal reporting from Mac OS X 10.6 onward, which is why a Mac and a Windows PC can show two different “available space” numbers for the exact same drive.
Worked Example 1: Why a “1TB” Drive Shows Only 931GB
A drive labelled 1TB genuinely contains 1,000,000,000,000 bytes — the manufacturer isn’t shortchanging you. The confusion happens only when an operating system measures that same byte count using 1,024-based units but still calls the result “GB.”
The math:
1,000,000,000,000 bytes ÷ 1,073,741,824 (bytes in 1 GiB) = 931.32
So the OS reports “931GB” — really 931 GiB — for a drive that was always, correctly, exactly 1TB. Run the same division for any capacity and you’ll land within a rounding error of what your own device shows:
- 256GB drive → 256,000,000,000 ÷ 1,073,741,824 ≈ 238.4 GiB, shown as “238GB”
- 512GB drive → 512,000,000,000 ÷ 1,073,741,824 ≈ 476.8 GiB, shown as “477GB”
- 2TB drive → 2,000,000,000,000 ÷ 1,073,741,824 ≈ 1,862.6 GiB, shown as “1.82TB”
Every single one of these is the identical number of bytes measured two different ways — nothing has actually gone missing.
Worked Example 2: Why Your 128GB Phone Never Has 128GB Free
A new 128GB phone loses space for two separate, stacking reasons, and it helps to tell them apart:
Reason 1 — the same binary/decimal gap as above:
128,000,000,000 bytes ÷ 1,073,741,824 = 119.2 GiB. Just from unit conversion, “128GB” becomes “119GB” the moment the phone’s settings screen measures it in GiB and labels it GB — before the operating system itself takes a single megabyte.
Reason 2 — the operating system, preinstalled apps, and “other” system data:
On top of that conversion gap, the OS itself, preloaded apps (camera, messages, browser, carrier apps), and reserved system partitions typically use several more gigabytes — the exact amount varies by brand, Android/iOS version, and even carrier, so there’s no single universal number, but industry reporting has repeatedly found real-world 128GB phones landing anywhere from roughly 112GB to 122GB of genuinely usable space once both effects are combined.
So when a 128GB phone shows “118GB available” out of the box, roughly 9GB of that gap is pure unit conversion (not recoverable, ever) and the rest is software you could theoretically free up by uninstalling preloaded apps.
Worked Example 3: Why a “100 Mbps” Plan Downloads at ~12.5MB/s, Not 100MB/s
This one isn’t a binary-vs-decimal issue at all — it’s a completely different unit hiding behind a similar-looking abbreviation. Internet and network speeds are measured in bits per second (a holdover from early modem and telecom engineering), while every file size, download manager, and storage spec is measured in bytes. One byte always equals 8 bits, so converting a connection speed into the download speed you’ll actually see just means dividing by 8.
The math:
100 Mbps ÷ 8 = 12.5 MB/s (megabytes per second) — the realistic best-case download speed on a “100 Mbps” plan, before any further loss from Wi-Fi overhead, server limits, or network congestion.
Worked through for a real download: a 2GB (2,000MB) movie file at a true, uncontended 12.5MB/s takes 2,000 ÷ 12.5 = 160 seconds, or about 2 minutes 40 seconds. At a smaller 50 Mbps plan (50 ÷ 8 = 6.25MB/s), the same file takes 2,000 ÷ 6.25 = 320 seconds, about 5 minutes 20 seconds — exactly double, since the speed is exactly half.
| Plan speed | Real download speed | 2GB file download time |
|---|---|---|
| 30 Mbps | 3.75 MB/s | ~8 min 53 sec |
| 50 Mbps | 6.25 MB/s | ~5 min 20 sec |
| 100 Mbps | 12.5 MB/s | ~2 min 40 sec |
| 300 Mbps | 37.5 MB/s | ~53 sec |
| 1 Gbps (1,000 Mbps) | 125 MB/s | ~16 sec |
Fibre providers in India almost always advertise the bigger bits-based number because it looks more impressive on a hoarding — a genuine 100 Mbps connection and a genuine 12.5 MB/s download speed are the same connection, just two different units describing it.
How to Convert Between Any Two Units Yourself
You don’t need a dedicated tool for most of these conversions — two formulas cover almost everything:
Decimal storage units (what’s on the spec sheet): move the decimal point 3 places for every step between B → KB → MB → GB → TB → PB. To go from a bigger unit to a smaller one, multiply by 1,000 per step; to go smaller to bigger, divide by 1,000 per step.
Binary units (what your OS actually measures internally): same idea, but each step is ×1,024 or ÷1,024 instead of 1,000. This is the number to use if you want to know exactly what a device will display, rather than what’s printed on the box.
Bits to bytes (connection speeds only): divide a bits-per-second figure by 8 to get bytes-per-second, and keep the same metric prefix (Mbps ÷ 8 = MB/s, Gbps ÷ 8 = GB/s, Kbps ÷ 8 = KB/s).
For everyday length, weight, temperature, volume, area, or speed conversions that aren’t digital-storage-specific, CheckMatter’s Unit Converter tool handles the arithmetic instantly — it’s worth bookmarking alongside this guide for the units it does cover, even though digital storage units sit in a separate system from the physical ones on that tool.
Common Mistakes
Writing “mb” when you mean “MB.” A lowercase “b” means bits, uppercase “B” means bytes — “100mb” (100 megabits) and “100MB” (100 megabytes) differ by a factor of 8, and spec sheets are not always careful about the case.
Assuming a drive is “missing” space. As shown in Worked Example 1, a correctly-labelled 1TB drive showing “931GB” has lost nothing — it’s the same byte count read through a different ruler.
Mixing decimal and binary mid-calculation. If you’re estimating how many 4MB photos fit on a drive, use the same base (1,000 or 1,024) for both the photo size and the drive size, or your estimate will drift by single-digit percentages for every unit involved.
Expecting your download manager to hit the plan’s advertised number. A “100 Mbps” plan was never going to show “100” in a download manager that reports MB/s — 12.5 is the correct, honest number for that exact same connection.
Frequently Asked Questions
Is 1 GB equal to 1,000 MB or 1,024 MB?
Both exist and both are “correct” depending on context. Officially (SI/IEC, and every storage manufacturer), 1 GB = 1,000 MB exactly. Many operating systems historically measured and displayed 1,024 MB while still calling it “GB” — that quantity’s correct name is 1 GiB, not 1 GB, though the mislabelling is extremely common.
Why do manufacturers use 1,000 instead of 1,024?
Decimal (1,000-based) units follow the same international SI system used for every other unit of measurement — kilometres, kilograms, kilowatts — so “kilo” always means exactly one thousand. It also happens to produce a slightly larger, more marketable number than the binary equivalent for the same physical capacity.
How many MB is a GB exactly?
Exactly 1,000 MB in decimal (manufacturer) terms, or exactly 1,024 MiB in binary (what your OS often measures) terms. Which one matters depends on whether you’re reading a product label or your device’s own storage screen.
Why is internet speed measured in Mbps instead of MB/s?
It’s a holdover from early telecom and modem engineering, where signal speed was naturally measured in bits. Using megabits also produces a number 8 times larger than the equivalent megabytes figure, which providers have little incentive to change.
How many photos or songs fit in a given amount of storage?
Divide the storage size by the average file size, using the same unit base throughout. A 64GB card (64,000MB in decimal terms) holding 4MB photos fits roughly 64,000 ÷ 4 = 16,000 photos; the same card holding 5MB songs fits roughly 12,800 songs. Treat these as rough planning estimates — real file sizes vary by resolution, format, and compression.
What’s the difference between a bit and a byte?
A bit is the smallest unit of digital data — a single 0 or 1. A byte is a group of 8 bits, and is the standard unit for measuring file sizes and storage, while bits are the standard unit for measuring data-transfer speeds.
Once you know which “ruler” a number is using — decimal or binary for storage, bits or bytes for speed — every one of these mismatches turns from a mystery into a two-line calculation.