Conflux's consensus is hybrid: Proof-of-Work for block production and ordering, Proof-of-Stake for finality checkpoints. PoW is still active. The September 2025 Conflux 3.0 hardfork — activated at epoch 129,680,000 (around September 1, 2025) — tuned both layers but did not eliminate PoW. If you mine CFX, mining still works. If you stake CFX, staking still works. The two systems complement each other rather than replacing each other.
How Conflux's hybrid consensus actually works
Most blockchains pick one consensus mechanism. Conflux runs two simultaneously, with each handling a different responsibility.
PoW (Proof-of-Work) handles block production. GPU miners produce candidate blocks using the Octopus algorithm — Conflux's PoW-of-choice. PoW gives Conflux censorship resistance: no central authority decides which transactions get included; whoever finds a valid block first proposes it. The Tree-Graph consensus algorithm allows multiple miners to produce parallel blocks and merges them into a directed acyclic graph, which is what gives Conflux its scalability.
PoS (Proof-of-Stake) handles finality. Stakers lock CFX into the protocol and vote on finality checkpoints — agreed-upon points in the chain history beyond which transactions cannot be reverted. PoS provides settlement-finality guarantees against long-range reorganization attacks that pure PoW chains are vulnerable to.
The combined system gets both properties: PoW's permissionless block production plus PoS's finality guarantees. Trade-offs: more computational complexity in the protocol; CFX issuance to both miners and stakers; staking participation matters for finality liveness.
What changed in Conflux 3.0 (September 2025)
The Conflux 3.0 hardfork activated at epoch 129,680,000 (around September 1, 2025). It included CIP-150 through CIP-165 — sixteen Conflux Improvement Proposals covering execution optimizations, EVM-compatibility improvements, and security tuning.
For PoW miners, the most relevant change was minimal — block-production rules and reward structure remain effectively unchanged. Conflux 3.0 was not a "merge to PoS" event in the way Ethereum's was.
For PoS stakers, the most relevant change was CIP-156: the slashing penalty for malicious validator behavior changed from permanent stake forfeiture to a temporary 6-month stake lock. This made validator participation lower-risk; misbehaving validators get their stake frozen rather than destroyed.
For DApp developers, CIP-7702 (set code for EOAs in eSpace) and CIP-165 (BLS12-381 precompiles) added Ethereum-compatible features that were missing from prior Conflux eSpace.
For CFX holders generally, the headline number was the network's increased designed throughput capacity — up to 15,000 transactions per second.
What this means for you
If you mine CFX with GPUs: mining is still a viable activity. The Octopus algorithm continues to produce blocks. Major mining pools (F2Pool, Woolypooly, Herominers, Nanopool) still support Conflux. Block rewards continue to be paid to PoW miners. No action required for the upgrade itself; just keep your mining software updated to the latest miner version compatible with the Conflux 3.0 protocol.
If you stake CFX: staking continues working. The 4% annualized rate is set by protocol parameters and was not directly changed by 3.0. The CIP-156 slashing change reduces risk for validators, which may indirectly affect staking-pool yield as more participants enter. If you stake through Conflux Wallet or any other compatible wallet, no action required.
If you hold CFX without mining or staking: no action required. The hardfork was non-disruptive for users. Your CFX balance, transactions, and addresses all remain valid. The Conflux Network experienced no halts during the transition.
If you operate a Conflux node: upgrade to v3.0.1 (the production-ready release) or later. Older versions will fall out of consensus once the hardfork activates. Most users don't run nodes; this matters only for infrastructure operators.
Why hybrid instead of pure PoS
Currently, no full PoS transition is scheduled. Conflux's design treats hybrid PoW + PoS as the long-term target, not a transition state. The Conflux Foundation has not announced an "Ethereum-style merge" roadmap. If a full PoS transition were to happen in the future, it would require a separate hardfork with months of community discussion. Any such proposal would emerge through the CIP process. As of April 2026, no such CIP has been proposed.
What you can do today
For most users, the answer is "nothing." Conflux 3.0 was a non-disruptive upgrade. Your wallet works the same way. Your CFX balance is unaffected.
For users actively staking, you may want to verify the current APY in your wallet (still ~4% annualized at the time of writing) and consider whether the lower-risk slashing rules change your validator-vs-delegator preference. For users actively mining, verify your mining software is current and your pool is still on the latest Conflux 3.0 protocol.
For users curious about the technical changes, the Conflux developer documentation has a complete CIP-by-CIP breakdown at doc.confluxnetwork.org/docs/general/hardforks/v3.0.
Sources
- Conflux Forum: v3.0.1 Hardfork Upgrade Completed (August 31, 2025)
- doc.confluxnetwork.org: v3.0 hardfork specification
- Crypto News Navigator: Conflux 3.0 throughput analysis (April 2026)