ContentAddressable

Content-addressable refers to systems that retrieve data by its content rather than by a location or traditional address. In these systems the address or key is derived from a hash of the data, typically a cryptographic hash, and the data object is stored and retrieved by this hash. Because the address encodes the content, identical data can be stored once and retrieved from any copy, enabling strong deduplication and integrity verification. If the content changes, the hash changes and a new address is produced, leaving the original object effectively immutable.

Content-addressable storage (CAS) is widely used in version control and distributed storage. In Git, for example, file contents, directories, and commits are stored as objects addressed by their hash. In distributed file systems such as IPFS, content is addressed via Merkle DAGs and retrieved by hash, supporting content-based routing and resilience to node availability.

Content-addressable memory (CAM) is a hardware approach in which memory can be accessed by the data value rather than by a numeric address, enabling parallel search and fast lookups. CAM is used in high-speed caches, pattern matching, and some network equipment, though it is more expensive and specialized than conventional RAM.

Advantages of content-addressable designs include data integrity verification, automatic deduplication, and simplified synchronization in distributed environments, as well as immutable data history. Drawbacks include the overhead of computing hashes, potential performance costs, and the need to manage content-based routing and persistence. Hash collisions are theoretically possible but typically negligible in practice with strong hash functions such as SHA-256.