Cardano just mapped out its biggest network change in years. The Cardano Dijkstra upgrade splits into two hard forks, and both target the same problem: transactions confirm too slowly, and the chain can’t yet handle heavy traffic. Intersect, Cardano’s governance body, published the roadmap on August 17. It spells out exactly how the network gets faster, without weakening the security guarantees that keep it running.
What Leios Changes First
Phase one lands in the fourth quarter of 2026 as Ouroboros Leios. It introduces something Cardano calls Endorser Blocks. These blocks let stake-based committees certify extra transactions alongside the main chain. Instead of forcing every transaction through the same narrow pipe, Leios opens a parallel lane. The chain processes more activity without slowing down the blocks it already produces.
Leios doesn’t travel alone. The same phase bundles four other improvements: nested transactions, guard scripts, upgraded account addresses, and simpler staking reward withdrawals. Developers already get to test the changes — Cardano launched a public testnet called Musashi Dojo back in June.
The Cardano Dijkstra Upgrade’s Second Act: Peras
Phase two arrives in the second quarter of 2027. Ouroboros Peras adds a voting layer that lets committees of stake pool operators weigh in on the chain’s most recent blocks. Right now, Cardano confirms a transaction by waiting for enough new blocks to stack on top of it. Peras lets operators vote a recent block “final” faster than that wait would normally allow. Users notice the difference immediately: payments and app transactions settle sooner.
Node Diversity Backs It Up
Speed means little if one piece of software controls the whole network. Cardano is pairing Dijkstra with Amaru, a node client written in Rust instead of the original Haskell. Amaru hits its next production milestones on September 30, October 29, and November 26. Running multiple independent clients protects the chain if a bug ever hits just one of them.
A Vote Decides the Fine Print
None of this ships automatically. Cardano’s delegated representatives, stake pool operators, and Constitutional Committee all have to approve a constitutional amendment that adds Dijkstra’s technical parameters. Intersect plans to submit that amendment by September 11. The vote won’t touch governance rules themselves — it only locks in the numbers Leios and Peras need to run.
Why the Cardano Dijkstra Upgrade Matters
Cardano isn’t alone in chasing speed this month. Solana just cut its own slot time down to 350 milliseconds, and the pressure across every major chain is the same: handle more people, confirm transactions faster, and don’t cut corners on security. Dijkstra shows Cardano can add real throughput and quicker finality through its own consensus research, rather than borrowing a shortcut other chains later have to walk back. If Leios and Peras land on schedule, Cardano moves from a network built for patience to one built for everyday use — payments, apps, and institutions that can’t wait minutes for a transaction to settle.