Qubic Consensus: UPoW, Quorum and the 676 Computor System

Related Guides in This Series

-> Quantum Proof Crypto: What It Means and Where Qubic Stands

-> Qubic Price: Token Economics and Market Dynamics

-> Qubic Games: Interactive Applications on the Network

How Qubic's Consensus Works

According to the Qubic documentation, Qubic was founded by Sergey Ivancheglo, creator of NXT and co-founder of IOTA. The network uses a Quorum-Based Computation (QBC) consensus model built around exactly 676 specialized validators called Computors. This architecture is fundamentally different from both traditional Proof of Work and Proof of Stake systems.

In a Proof of Work system like Bitcoin, any miner can propose blocks and the longest chain wins. In a Proof of Stake system, validators are chosen based on staked holdings. Qubic's model uses a fixed set of 676 Computors selected by performance each epoch, requiring a strict supermajority to agree before any transaction reaches finality.

The 676 Computor Architecture

Fixed Validator Set

The Qubic network operates with exactly 676 Computors at all times. This fixed number creates a defined and predictable network topology. Computors are responsible for executing smart contracts, securing the network, and processing transactions. New participants enter the network as Candidates, contributing to AI training tasks without receiving direct compensation until they earn a top-676 ranking.

Supermajority Consensus Threshold

Finality on Qubic requires agreement from at least 451 of the 676 Computors, representing approximately 67% of the total validator set. This two-thirds supermajority is a Byzantine Fault Tolerant threshold, meaning the network remains secure and operational even if up to one-third of Computors are offline, faulty, or acting maliciously.

Once 451 Computors agree on a transaction, it achieves instant and irreversible finality. There is no probabilistic confirmation period as in Proof of Work systems where you wait for multiple block confirmations. Settlement is final from the moment quorum is reached.

Epoch-Based Operations

According to the tokenomics documentation, each epoch spans exactly seven days. At the end of each epoch, Computors are re-ranked by the performance of their AI miners. The top 676 performers retain or attain Computor status for the next epoch, while lower-ranked participants become Candidates. This competitive structure creates continuous incentives for Computors to maintain high performance.

Useful Proof of Work (UPoW)

Beyond Traditional Mining

According to the UPoW documentation, Qubic's Useful Proof of Work system transforms the energy spent in mining from arbitrary hash calculations into something productive. Rather than directing computational power toward solving mathematical puzzles that serve only to secure the chain, UPoW directs that energy toward AI training tasks.

The network presents AI models with complex training challenges, such as processing large datasets and training machine learning models on specific problems. Computors, backed by AI miners, compete to solve these tasks. Their performance in solving these tasks determines their epoch ranking, which determines whether they retain Computor status.

The Role of Aigarth

The Qubic About page describes Aigarth as more than AI. Powered by a global network of miners, Aigarth evolves by creating billions of artificial neural networks, building the future of AI in a decentralized and open way. The UPoW mining competition drives this AI development by directing global compute capacity toward Aigarth's training requirements each epoch.

Dual-Purpose Security Budget

Traditional Proof of Work uses energy to produce cryptographic proofs that serve only to secure the ledger. With UPoW, the same energy expenditure serves two purposes simultaneously: establishing Computor rankings that determine network consensus, and advancing real-world AI research. This design addresses one of the core criticisms of Proof of Work, that mining energy has no external value beyond network security.

Smart Contract Governance

According to the Qubic documentation, smart contracts on Qubic are written in C++ and executed directly on bare metal hardware rather than through a virtual machine. This bare metal execution eliminates the overhead of VM interpretation and allows full utilization of hardware capabilities.

Before any smart contract can launch on the Qubic network, it must pass a quorum vote by Computors. This governance mechanism prevents unauthorized or malicious contracts from deploying without distributed community approval. The same quorum that secures transactions also governs which smart contracts are permitted to operate.

Smart contract execution requires a commission, which Computors vote by quorum to determine. This commission is burned permanently rather than distributed to validators, contributing to the deflationary economic model described in the tokenomics documentation.

Smart contracts on Qubic launch through an IPO model: once approved by Computor quorum, shares in the contract are auctioned to the community, enabling passive income for shareholders and novel economic structures.

Token Economics and Consensus Incentives

The complete tokenomics framework is documented at docs.qubic.org/learn/tokenomics.

• Supply cap: 200 trillion QUBIC, reduced from 1,000 trillion by community vote

• Epoch duration: seven days

• Current emissions (post-Epoch 175 halving, August 2025): 450 billion QUBIC per epoch

• Each epoch's QUBIC is allocated across Computors, the Computor Controlled Fund (CCF), QEarn, and burns

• A Computor operating at maximum efficiency can receive approximately 1.479 billion QUBIC per epoch (at pre-halving emission rates of 1 trillion / 676 Computors)

• Transfers between wallets are completely feeless

• Smart contract commissions are determined by Computor quorum vote and burned, not paid to validators

The first halving occurred at Epoch 175 in August 2025, reducing per-epoch emissions from 850 billion to 450 billion QUBIC. A Supply Watcher Smart Contract dynamically adjusts emissions based on real-time supply data.

How Qubic Consensus Compares

The key distinctions between Qubic's consensus model and alternatives:

• Versus Bitcoin PoW: Qubic's consensus is not hash-racing. Finality is instant upon quorum agreement, not probabilistic. Energy is directed toward productive AI work, not arbitrary hashing.

• Versus Ethereum PoS: Qubic uses a fixed Computor set of 676 rather than a large and variable validator pool. Finality is instant through BFT quorum rather than a multi-step finalization gadget.

• Versus other BFT chains: The integration of UPoW AI training into Computor ranking creates a productive use for mining energy that most BFT systems lack entirely.

Key Reference Links

Introduction documentation: https://docs.qubic.org/overview/introduction/

UPoW documentation: https://docs.qubic.org/learn/upow/

Tokenomics documentation: https://docs.qubic.org/learn/tokenomics/

Smart contracts documentation: https://docs.qubic.org/learn/smart-contracts/

Qubic main website: https://qubic.org

Explorer: https://explorer.qubic.org

Qubic blog: https://qubic.org/blog-grid

Official Qubic Resources

qubic.org  |  docs.qubic.org  |  wallet.qubic.org  |  explorer.qubic.org