DNS and IP are two independent concepts.
In an IP network, you take a block of IPs, for example the private range 192.168.0.0/16, and divide it into smaller blocks, for example 192.168.1.0/24, 192.168.2.0/24 and so on. This divides the big network 192.168.0.0/16 into 256 independent networks in my example, and can be individually described with their own network address in a route for example, you can still collectively refer to them as the /16 network for aggregate routes for example. Thus, we have "subnetted" the larger network into subnetworks.
DNS has similar concepts, but it's entirely separate from IP. DNS is essentially a globally distributed hierarchical database, where different servers are authoritative for different zones. Zones can be defined as essentially anything, but commonly it's one subdomain. So in your example sea.microsoft.com. there would be at least 4 zones. First is the root zone, or the trailing . that contains the records over which servers are authoritative for each tld. Next is the .com zone, which contains records for the servers of each domain. Third is .microsoft, forth is sea. Normally, people define the first part of the DNS record that isn't a tld as the domain (in this case microsoft), and anything after that is a subdomain. But, the subdomain can contain anything, not only IP addresses, and DNS records usually contain only host IPs, for example the mail server IP addresses for sea.microsoft.com. As such, while you can define some sort of logic between your IP blocks and subdomains, there's not relation between the two.
The only exception is reverse lookup records for public IP addresses, but that's another topic entirely.