Consensus Mechanism

6.1Why Not Classic Proof-of-Work or Proof-of-Stake

Classic proof-of-work, as used by Bitcoin, has miners race to solve a computational puzzle, with the first valid solution winning the block and competing solutions discarded — a process that necessarily wastes most of the computation spent and requires waiting for multiple confirmations to be confident a block will not be reorganised.

Classic proof-of-stake, as used by most modern chains, replaces that computational race with a selection process weighted by how much currency a validator has staked, which solves the energy-waste problem but introduces a different one: a validator's influence is purely a function of capital, with no link to real-world energy or activity.

Pim Protocol's consensus mechanism, Proof of Entropy Minima (PoEM), takes a third path: it keeps a deterministic, single-winner selection process — like proof-of-work, but without the wasted competing computation — while linking the system's monetary policy to a verifiable measure of real energy cost, described fully in Section 7.

6.2Proof of Entropy Minima — Mechanics

Every candidate block is assigned a weight using a cryptographic hash function applied to its header, a nonce value, and a timestamp:

Weight = BLAKE3-XOF(Block Header, Nonce, Timestamp)

The candidate block with the lowest weight wins that slot. Because the hash function's output is effectively unpredictable until computed, and because every node computes the same weight from the same inputs, there is exactly one correct winner for each slot — no two honest nodes can disagree about which block should be selected, which is what eliminates the fork races inherent in classic proof-of-work. This selection step completes in roughly 0.01 milliseconds, and the probability of a genuine tie (two blocks computing equally low weights) falls to about 0.0007% even in a 10,000-node network.

6.3Supporting Mechanisms

Cross-Shard Merged Mining (CSMM) allows a single miner's hardware to contribute proof-of-work toward multiple shards in parallel, rather than dedicating separate hardware to each shard, cutting per-shard overhead by an estimated 30–40%.

The Deterministic Finality Protocol (DFP) handles transactions that touch more than one shard through a coordination process called Prime Shard Braiding, reaching full cross-shard finality in approximately 8 seconds.

The GHOST fork-choice rule serves as a backstop for the rare residual conflict — in practice fewer than 0.01% of all blocks — with an average resolution time under one second.

6.4Finality and Confirmation

The "owned-object fast path" reflects a design borrowed conceptually from object-model chains such as Sui: a transaction that only touches ledger objects owned exclusively by the sender — for example, a governance vote cast from a single voter's own staking record — does not need to wait for full network-wide consensus, because there is no possibility of a conflicting claim on that object from anyone else.