Data Size Converter

The Hidden Confusion in Digital Storage

Every year, millions of consumers buy a "1 terabyte" hard drive and feel vaguely cheated when their computer reports it as 931 gigabytes. Nobody stole 69 gigabytes. The discrepancy is entirely a matter of definitions — and understanding it requires knowing the difference between two competing standards that the technology industry has stubbornly refused to fully reconcile.

Storage manufacturers define their capacities in SI (International System) units, where 1 kilobyte = exactly 1,000 bytes, 1 megabyte = 1,000,000 bytes, and 1 terabyte = exactly 1,000,000,000,000 bytes. This is mathematically clean and matches the metric system. However, computer engineers working with binary systems found it more natural to use powers of 2: 2¹⁰ = 1,024 bytes was called a "kilobyte," 2²⁰ = 1,048,576 bytes was called a "megabyte," and so on. Operating systems like Windows still report file sizes and storage capacity in these binary multiples, even though they label them "GB" rather than the technically correct "GiB." The result: a 1 TB drive (1 × 10¹² bytes) divided by 1,073,741,824 bytes/GiB equals approximately 931 GiB — which Windows labels as 931 GB.

To eliminate this ambiguity, the International Electrotechnical Commission (IEC) introduced a new set of prefixes in 1998 under the IEC 80000-13 standard: kibibyte (KiB) for 1,024 bytes, mebibyte (MiB) for 1,048,576 bytes, gibibyte (GiB) for 1,073,741,824 bytes, and so forth. Linux adopted these prefixes widely, and macOS switched to SI definitions (matching manufacturer labels) starting with macOS Snow Leopard in 2009, which is why Mac users typically see a 1 TB drive as approximately 1 TB — their OS uses the same definition as the manufacturer.

The explosive growth of data in the modern era has made these distinctions more significant than ever. According to IDC projections, the amount of data created, captured, copied, and consumed globally was on pace to reach approximately 120 zettabytes annually by 2025. One zettabyte is 10²¹ bytes — a trillion gigabytes. The entire early internet's indexed content in 2000 was estimated at around one exabyte (one billion gigabytes). We now create that much data every few hours.

Network speeds and storage sizes use different units, which causes widespread confusion. When your internet provider advertises "100 Mbps," that is 100 megabits per second — not megabytes. Since there are 8 bits in a byte, your maximum download speed in megabytes is 100 ÷ 8 = 12.5 MB/s. A 4 GB file download at 12.5 MB/s would take about 320 seconds — roughly five minutes. This is why downloading a large game feels slower than the advertised connection speed suggests.

Streaming services have become one of the largest consumers of bandwidth and storage. Netflix uses approximately 0.7 GB per hour for standard definition, 3 GB per hour for HD (1080p), and 7 GB per hour for Ultra HD 4K. At an average of 4.5 hours of TV watched daily per American household, Netflix and similar services account for a substantial fraction of global internet traffic — Sandvine estimated streaming video represented over 65% of downstream internet traffic in North America in recent years.

At the cutting edge of storage research, scientists are investigating DNA as an archival medium. DNA data storage encodes information in the four nucleotide bases (A, T, G, C), with each base representing approximately 2 bits. In theory, one gram of DNA can hold approximately 215 petabytes of information — roughly 215 million gigabytes. The challenge is the cost and speed of synthesizing and reading DNA sequences, which remain orders of magnitude slower and more expensive than conventional storage. But researchers at Microsoft and the University of Washington have demonstrated reading and writing of small amounts of data in DNA, and the longevity of the medium — DNA can survive thousands of years under the right conditions — makes it compelling for deep archival storage.

Compression is the other major dimension of storage efficiency. Lossless compression (ZIP, FLAC audio, PNG images) reduces file size without discarding any information — every bit can be perfectly reconstructed. Lossy compression (MP3, JPEG, H.264 video) achieves much higher compression ratios by discarding information the human perceptual system is unlikely to notice. A 40 MB raw WAV audio file typically compresses to 3–5 MB as an MP3 at 192 kbps — a 90% size reduction at a quality level most listeners cannot distinguish from the original. Understanding what level of compression is appropriate for your use case is often more practically important than understanding the unit conversions themselves.