IP Address Notations: Decimal, Binary, Hex, and the 32-bit Integer Form

One 32-bit number, several ways to write it

An IPv4 address is fundamentally a single 32-bit number — every notation you'll encounter is just a different way of writing that same value, chosen for whichever context makes it most convenient to read or process.

Dotted-decimal: 192.168.1.1
Binary:         11000000.10101000.00000001.00000001
Hex:            C0.A8.01.01
Packed integer: 3232235777

Dotted-decimal: the human-readable default

Splitting the 32 bits into four 8-bit (0–255) octets, each written in decimal and separated by dots, is the format almost everyone recognizes — chosen specifically because four small decimal numbers are far easier for a person to read, remember, and type correctly than a 32-bit binary string or a 10-digit integer.

Binary: what a subnet mask calculation actually operates on

Binary is rarely how addresses are stored or displayed in day-to-day tools, but it's the representation that makes subnetting logic transparent — a CIDR prefix (see CIDR Subnetting Explained) is literally a statement about how many leading bits are fixed, which is far more direct to reason about when the address is written in binary rather than decimal.

Hex: compact, and common in low-level networking tools

Hexadecimal packs the same 8 bits per octet into exactly 2 characters (C0 instead of 192), which shows up in packet capture tools, some router configuration interfaces, and IPv6 addresses (which use hex groups natively, unlike IPv4).

The packed 32-bit integer: what databases actually store

This is the form most people don't encounter directly but that matters a great deal behind the scenes: many databases, geolocation datasets, and IP-range lookup systems store addresses as a single 32-bit integer rather than four separate octets, because a single integer is both more storage-efficient and trivially comparable — checking whether an address falls within a range becomes a simple numeric comparison (start <= ip <= end) instead of comparing four separate octet fields with awkward carry logic.

192.168.1.1
= 192×256³ + 168×256² + 1×256¹ + 1×256⁰
= 3232235777

Common mistakes

• Assuming dotted-decimal is the "real" underlying format. It's simply the most human-readable of several equivalent representations of the same 32-bit value.
• Manually converting octets to a packed integer incorrectly. Each octet's positional weight is a power of 256, not 10 or 16 — a common arithmetic slip when converting by hand.
• Forgetting hex requires exactly 2 characters per octet. A single-digit hex value needs a leading zero (0A, not A) to represent a full byte correctly in most conventions.

FAQ

Why do databases often store IP addresses as a single integer instead of the usual dotted format?
A packed integer is more storage-efficient and makes range comparisons (is this address between X and Y) a simple numeric operation, rather than requiring per-octet comparison logic.

Is binary notation ever actually used in networking tools, or is it purely educational?
It's genuinely used for subnetting logic — a CIDR prefix directly describes a count of fixed bits, which is most transparent to verify when the address is written out in binary.

How many hex characters represent one IPv4 octet?
Exactly 2 — an 8-bit value (0–255) maps to exactly two hexadecimal digits (00–FF).

Convert any IPv4 address between dotted-decimal, binary, hex, and packed integer instantly with the IP Address Converter — entirely in your browser.