Data Storage Units: Bits, Bytes, and Beyond

Introduction

When you buy a 1 TB hard drive, your operating system reports it as 931 GB. When your internet provider advertises 100 Mbps, a 1 GB file takes more than 80 seconds to download. These discrepancies are not marketing tricks — they stem from a genuine conflict between two competing systems for naming data storage units: the binary system used by computers and the decimal system used by drive manufacturers and network engineers.

Understanding the difference between KB and KiB, GB and GiB, and Mbps and MBps eliminates confusion and helps you make accurate calculations.

The Basics: Bits and Bytes

Bit

A bit is the smallest unit of digital information. It has exactly two possible values: 0 or 1. All digital data — images, text, video, executable code — is ultimately stored as sequences of bits. The word "bit" is a contraction of "binary digit."

Byte

A byte is 8 bits. This grouping became standard in the 1960s with the IBM System/360. One byte can represent 256 distinct values (2⁸ = 256), enough to encode a single character in ASCII.

All modern storage and memory is measured in bytes or multiples of bytes. Networking speed is typically measured in bits per second.

This distinction matters constantly: a 100 Mbps (megabits per second) connection transfers 12.5 MB (megabytes) per second. Divide by 8 to convert bits to bytes.

The Two Prefix Systems

Decimal Prefixes (SI)

The International System of Units defines prefixes based on powers of 10:

Hard drive manufacturers, storage device makers, and network engineers use SI decimal prefixes. When a manufacturer says "1 TB," they mean exactly 1,000,000,000,000 bytes.

Binary Prefixes (IEC)

Computers allocate memory and file systems in powers of 2. Early in computing history, engineers used the same SI prefixes (KB, MB, GB) but meant powers of 2 because 2¹⁰ = 1,024 is close to 1,000. The International Electrotechnical Commission (IEC) formalized distinct binary prefixes in 1998:

Windows File Explorer reports sizes in binary GiB but labels them "GB." macOS switched to reporting in decimal GB in 2009. Linux tools often report in either system depending on the command.

Why Your Drive Shows Less Space

A 1 TB drive contains exactly 1,000,000,000,000 bytes (decimal TB). Windows divides this by 1,073,741,824 (GiB) instead of 1,000,000,000 (GB) and calls the result "GB":

1,000,000,000,000 ÷ 1,073,741,824 = 931.32 GiB

Windows displays this as "931 GB." The drive is not defective — the OS is reporting binary GiB using the decimal GB label. The discrepancy grows with drive size:

Use the data storage converter to calculate exact equivalents.

Networking: Bits vs. Bytes Per Second

Network speeds are almost universally measured in bits per second with SI prefixes:

To find the file transfer rate in MB/s, divide Mbps by 8: - 100 Mbps ÷ 8 = 12.5 MB/s - 1 Gbps ÷ 8 = 125 MB/s - 10 Gbps ÷ 8 = 1,250 MB/s

A 4K movie file of 50 GB on a 100 Mbps connection takes: 50,000 MB ÷ 12.5 MB/s = 4,000 seconds ≈ 66.7 minutes

On a 1 Gbps fiber connection: 50,000 MB ÷ 125 MB/s = 400 seconds ≈ 6.7 minutes.

Real-World File Sizes

Understanding storage units is more intuitive with concrete examples:

Historical Growth in Storage

The growth of storage density over the past 70 years is extraordinary:

• 1956: IBM 350 (first commercial hard drive) — 5 MB, required a machine the size of a refrigerator, weighed 1 ton
• 1980: First 5.25" hard drives — 5–10 MB, desktop-sized
• 1991: IBM 0663 — 1 GB, cost approximately $3,000
• 2000: Consumer hard drives reached 30–80 GB
• 2010: First consumer 2 TB drives
• 2024: Consumer hard drives reach 28 TB; SSDs reach 8 TB in 2.5" form factor

Areal density (bits per square inch) has doubled roughly every two years — an analog of Moore's Law for storage. Price per gigabyte has fallen from approximately $300,000 per GB in 1956 to less than $0.02 per GB for spinning disk in 2024.

Flash Storage and SSDs

NAND flash memory, the basis of SSDs, USB drives, and SD cards, stores data in floating-gate transistors. Modern 3D NAND stacks 200+ layers vertically and uses:

• SLC (Single-Level Cell): 1 bit per cell — fastest, most durable, most expensive
• MLC (Multi-Level Cell): 2 bits per cell
• TLC (Triple-Level Cell): 3 bits per cell — common in consumer SSDs
• QLC (Quad-Level Cell): 4 bits per cell — highest density, lower write endurance

A consumer TLC SSD typically has a rated endurance of 300–600 TBW (terabytes written). At 20 GB/day of writes, that is 40–80 years of use — more than the likely lifespan of the device for other reasons.

Key Conversion Reference

Conclusion

Data storage units involve two parallel naming systems that use the same words to mean different things. Drive manufacturers and network engineers use decimal SI prefixes; operating systems and memory use binary prefixes. The IEC binary prefix standard (KiB, MiB, GiB) was designed to resolve this ambiguity, but adoption remains inconsistent. When precision matters — pricing storage, calculating download times, or planning disk partitions — always confirm whether you are working in decimal or binary units. Use the data storage converter to move between any combination instantly.