Input Output Targets 3–5x Capacity for Cardano (ADA) in Staged Leios Rollout

Input Output Stages Leios Behind Dijkstra

Input Output has set out a staged path for Cardano (ADA) to scale through its Leios upgrade, and most of the promised capacity sits behind steps the network has not yet taken. In its Oct. 20 update, the developer targets an initial three to five times today's transaction capacity, with a later 10 to 15 times available only after separate governance votes. The plan separates the network upgrade from the steps needed to activate Leios and permit more traffic. Leios arrives with the Dijkstra hard fork, after which stake pool operators must register new BLS voting keys covering an estimated 90–95% of delegated staking before the upgrade can operate. No activation date for Dijkstra is fixed, and the stated gains stay subject to mainnet load results and community approval. Against the current Praos ceiling of roughly 4.5 kilobytes of transaction data per second, the initial target works out to about 13.5–22.5 TxkB/s, and the later goal to 45–67.5 TxkB/s. Those figures measure data throughput, not transactions per second, since transaction size decides how many fit. “The hard fork makes the protocol available,” Input Output wrote. The rollout covers capacity, not the Cardano price; as a proof-of-stake chain, Cardano (ADA) routes activation through pool operators and delegated representatives. The sequencing makes Cardano scaling a coordination problem as much as a software one. Dijkstra preparation has moved through separate software releases, with September's node 11.1.1 addressing operational issues ahead of versions carrying the new era's features, and Intersect's two-phase roadmap places Linear Leios first, with Peras as a separate upgrade in the second quarter of 2027.

Mechanically, the plan leans on BLS signatures, which let hundreds of pool votes compress into a single compact certificate. Leios needs votes representing at least 75% of active stake to certify an endorser block, the record that carries references to transactions awaiting processing, so registration must sit comfortably above that line to absorb pools that are offline or slow to respond. Input Output estimates participation from roughly the top 500–900 pools, out of about 2,700 registered on the live network, and each registration takes effect at the next epoch boundary. Pools without a registered key can keep producing Praos blocks but cannot vote on endorser blocks. Governance thresholds differ by stage: the constitutional amendment covering the new protocol parameters needs support from delegated representatives holding 75% of stake plus two-thirds of the Constitutional Committee, the hard fork needs 60% of DRep stake, 51% of stake pool operator stake and two-thirds of the committee, and later throughput changes need 67% of DRep stake, 51% of SPO stake and two-thirds of the committee. The preparation follows the July 18 van Rossem hard fork, which moved the chain to Protocol Version 11. Stress data comes from testing: during a 41-day Earth phase on the MusashiNet testnet, the team's Aug. 20 engineering account recorded a peak of 26.8 TxkB/s against a 4.51 TxkB/s Praos ceiling, a ratio of about 5.94x, across more than 127,000 blocks and 63 participating pools. The same account documented memory leaks, chain forks, stalled synchronization and a network halt caused by an invalid certificate handled as a fatal error, all of which the team said it corrected. Traffic in that test came from synthetic load generation and operators, so it describes testnet performance rather than customer demand. Once live, transactions routed through endorser blocks are expected to reach the ledger in roughly 45–60 seconds.

The 50–70 TxkB/s Break-Even Line

The economics give the timeline its stakes. Input Output's update notes that more than 85% of epoch rewards currently come from Cardano's Reserve, and cites a CIP-164 estimate that fee income would need sustained traffic of roughly 50–70 TxkB/s to break even, a band only the later 45–67.5 TxkB/s target reaches. Demand context matters as well, and recent adoption runs beyond finance, with Georgia wine producers logging 100,000 bottles on the chain. Operators preparing for the registration window can start with choosing a Cardano stake pool. With Cardano (ADA) trading 93% below its $3.10 peak, capacity only pays if traffic arrives, and current levels sit in our Cardano technical analysis. Until a Dijkstra date is fixed, the controlled traffic bursts, conservative initial settings and constitutional guardrails on what governance can change without another fork are the mitigation in place.