Directed Acyclic Graph (DAG)

A Directed Acyclic Graph (DAG) is a mathematical structure made of nodes connected by one way edges that never loop back on themselves. In cryptocurrency, several networks use this shape instead of a chain of blocks to record transactions, letting a new entry attach to multiple earlier ones at once rather than waiting in a single sequential queue.

Because a DAG does not force every transaction through one linear chain, many can be validated in parallel, which is the main reason projects choose it: higher throughput and, in some designs, no separate miners or per transaction fees. IOTA's Tangle is the best known example: each new transaction references and approves two prior ones, so participants collectively secure the ledger simply by transacting on it. Hedera's Hashgraph is a related but distinct approach, where nodes "gossip about gossip", sharing transaction history with each other; its underlying data structure is also a DAG, and the network reports finality in a few seconds with throughput far above older blockchain designs. Both aim to solve the same bottleneck that strictly linear chains face as usage grows.

DAG based ledgers involve trade-offs. Many rely on newer, less battle tested consensus mechanisms, and some early implementations, including IOTA's original Tangle, used a coordinator node to prevent conflicting transactions, a design critics viewed as a form of centralization until it was later removed. As these networks continue to mature, DAGs remain one of several architectures, alongside sharding and layer-2 rollups, that developers use to push transaction speed and scalability beyond what a single-file blockchain can achieve.