Supernetting is subnetting in reverse

Subnetting fixes more bits to make a block smaller. Supernetting does the opposite: it fixes fewer bits so a single prefix covers several smaller ones. Take two adjacent blocks, 192.168.0.0/24 and 192.168.1.0/24. Together they span 192.168.0.0 through 192.168.1.255, which is exactly the range of 192.168.0.0/23. Drop one bit of prefix, from /24 to /23, and one route now describes both blocks.

This is the engine behind route aggregation, also called summarization: replacing a list of specific prefixes with one shorter prefix that covers them all.

The rule for combining two blocks

Two blocks merge into one only when they form a sibling pair: same size, adjacent, and with the lower block sitting on the boundary of the combined block. 192.168.0.0/24 and 192.168.1.0/24 qualify, because 192.168.0.0 is a valid /23 network. But 192.168.1.0/24 and 192.168.2.0/24 do not combine, even though they look adjacent. 192.168.1.0 is not a /23 boundary (the /23 containing it starts at 192.168.0.0), so there is no single /23 that holds exactly those two blocks and nothing else.

Alignment, not mere adjacency, is what makes aggregation work. It is the same power-of-two boundary rule that governs subnetting, read from the other direction.

Exact aggregation versus a single supernet

When your prefixes are not all contiguous, there are two different answers, and they are not the same thing.

The minimal covering set is the shortest list of prefixes that covers exactly your inputs and nothing more. Given 10.0.0.0/24 and 10.0.2.0/24, there is a gap at 10.0.1.0/24, so they cannot be merged: the minimal set is still the two /24 blocks.

The single supernet is the one smallest prefix that contains every input, even if it pulls in addresses you did not list. For the same two blocks that is 10.0.0.0/22, which spans 10.0.0.0 to 10.0.3.255. It covers both inputs but also adds 10.0.1.0/24 and 10.0.3.0/24, 512 extra addresses you may not own.

Advertise the single supernet only when you actually control the whole range. Otherwise you summarize across address space that belongs to someone else.

Why it matters

Aggregation is what keeps routing tables small. One summary route in place of dozens of specifics means less memory on every router that hears it, and far less churn when a single component link flaps. The catch is that it only works if addresses were allocated contiguously and on aligned boundaries in the first place, which is why disciplined subnetting and VLSM pay off later as clean route summarization.

You can try both answers in the CIDR calculator. Its Supernet mode shows the minimal covering set and the single enclosing supernet, with the exact count of extra addresses the supernet would add.