addressspace

Address space is the range of addresses that a system component can reference. In computing, it usually refers to memory address spaces—the set of addresses a processor or memory subsystem can address—and may also refer to I/O address spaces in hardware. In networking, the term can describe the total pool of possible addresses in a given protocol, such as IP addresses.

Physical versus virtual: The physical address space corresponds to actual RAM addresses. The virtual address space is a layer of indirection provided by the operating system and the memory management unit (MMU), allowing processes to use addresses independent from actual hardware. The MMU translates virtual addresses to physical addresses, typically via page tables, perhaps with a translation lookaside buffer (TLB).

Size and structure: The size of an address space is determined by the width of the address representation and by architectural choices. A 32-bit system can address up to 4 GiB of physical RAM; 64-bit systems can address vastly larger spaces, though practical limits are determined by hardware and OS support. Virtual address spaces per process may be smaller or larger than physical memory; some systems provide 2–3 or more multiples of RAM.

Management considerations: Address space management includes allocation and protection, preventing one process from accessing another's memory. Techniques include paging, segmentation, and address space layout randomization for security. Fragmentation, TLB misses, and limited address spaces can affect performance and capacity.

IP address space: In networking, the IP address space defines the range of possible addresses. IPv4 provides about 4.29 billion addresses, while IPv6 provides a vastly larger pool, addressing exhaustion and routing considerations. Address allocation and subnetting organize the space for efficient routing and network administration.