/20 subnet question

[njnetwork]

Hello all,

I'm trying to answer the following question: 

What is the next subnet ID for the IP address: 172.29.48.0 /20

/20 block size is 4096 addresses (not subtracting 2 here, because we're not calculating hosts)

How do I get from 172.29.48.0 to the next network address of 172.29.52.0? The network blocks are incrementing by 4 in the third octet, but I don't understand why. I've scoured the internet trying to answer this question, and am not finding what I need. Chat GPT was also unhelpful. 

Any clarification would be greatly appreciated.

[MHM Cisco World]

[njnetwork]

Thank you for you response, but this is not helpful. I know how to use a subnet calculator. I'm trying to understand what calculation is used to increment subnets this large. I need the why, not the answer. 

[MHM Cisco World · Accepted solution]

4096 hosts 
each netework bit give you 255 hosts 
4096/255 host = 16 
that why you see the network-network interval is 16 

[Joseph W. Doherty · Accepted solution]

"How do I get from 172.29.48.0 to the next network address of 172.29.52.0?"

You don't, if you're using /20s.  The next network address, for a /20, would be 172.29.64.0.

172.29.52.0/20 is within 172.29.48.0/20, i.e. it's a host address.

BTW, another way to understand the situation, is "watch" the binary values.

A /20 divides the 32 bit IP address as: nnnnnnnn.nnnnnnnn.nnnnhhhh.hhhhhhhh.

In the third octet, the last network bit is the 4th bit, of 8.  I.e. binary 10000 = 16 decimal, which is why you see, in @MHM Cisco World first posting, the 3rd octet jumping by 16s.

"/20 block size is 4096 addresses (not subtracting 2 here, because we're not calculating hosts)"

Yes, that's correct, each /20 contains an IP address range of 4,096 because the largest host number, in binary, is 1111 1111 1111 or 4,065 decimal, plus counting zero, provides 4,096 numbers.

[njnetwork]

Thank you! This is super helpful.

[Joseph W. Doherty]

You're most welcome.

One of IPv4's mysteries, to me, is why IP addresses are represented as a dot-decimal.  Maybe because it was seen to go well with Classful addressing?

Certainly makes many struggle when you need to use ACLs and/or subnet/supernet.