Kinetiq Launches Elysium Mainnet With 100-Millisecond Blocks

By Nathan Brooks

Kinetiq launched Elysium’s public mainnet on Monday, Nov. 9, with 100-millisecond blocks and an estimated 300-millisecond user-visible transaction receipt time, according to Kinetiq. The Hyperliquid layer-2 gives applications a separate environment for processing transactions before settlement on HyperEVM.

Kinetiq is targeting 300 million gas units per second of execution. The block interval equals one-tenth of a second, while the estimated receipt time represents three-tenths of a second. Users pay transaction fees in HYPE.

What does Elysium change for Hyperliquid applications?

Elysium inserts a dedicated execution layer between Hyperliquid applications and HyperEVM settlement. Contracts deployed on Elysium can read HyperCore order books, prices, balances and positions. That access allows application logic to work with platform data without requiring transaction activity to return to the HyperEVM environment first.

For trading interfaces and other latency-sensitive applications, the relevant figure is the estimated 300-millisecond receipt time. That metric concerns transaction visibility, not final withdrawal. A user can receive an Elysium transaction receipt long before assets moving back to HyperEVM clear the network’s separate withdrawal process.

The 300-million-gas-per-second figure also requires careful interpretation. Kinetiq describes it as an execution target; the launch details do not establish sustained throughput under production load or specify the hardware configuration behind the number. Application teams evaluating capacity therefore have a stated design goal, not a disclosed stress-test result.

Fees and withdrawal mechanics

Users pay Elysium transaction fees in HYPE. Kinetiq said HYPE “can move from HyperEVM to Elysium on a one-to-one basis without a wrapped token.” For builders and treasury operators, this removes a wrapped representation from the stated transfer flow and keeps the fee asset aligned with the underlying Hyperliquid token.

Withdrawals back to HyperEVM follow a different path from ordinary Elysium transaction execution. They include a challenge period of about 5.5 days, equivalent to 132 hours. That window defines a material settlement constraint for applications that need capital back on HyperEVM, even though transaction receipts inside Elysium arrive in fractions of a second.

Launch infrastructure

Elysium runs on Arbitrum technology, while Conduit operates its infrastructure. At launch, one sequencer orders transactions, and one Conduit server provides data availability, meaning that transaction records remain accessible to the network. The disclosed setup therefore has one transaction-ordering point and one Conduit-operated server for data availability.

This architecture gives applications a clear initial dependency model. It also makes the sequencer’s ordering role and the availability of its designated server central operational dependencies. The launch configuration does not describe a second sequencer or additional data-availability servers, so teams building around Elysium should treat the disclosed topology as the relevant baseline rather than assume distributed operation that Kinetiq has not specified.

The separation between fast execution and slower withdrawals is also material for market structure. Elysium can order and expose application activity quickly, while capital returning to HyperEVM remains subject to the roughly 5.5-day challenge period. The two processes should not be treated as interchangeable forms of finality.

The four-week roadmap

Kinetiq plans an upgrade roughly four weeks after the Nov. 9 launch. It will add a native HyperCore market-data precompile and a fast-write channel. The first feature formalizes access to the order-book, price, balance and position data available to Elysium contracts as a protocol-level precompile. The second adds a dedicated channel for state updates, although Kinetiq has disclosed no separate performance target for it.

For engineering teams, the distinction is concrete: the precompile supplies a native contract routine for market data, while the fast-write channel supplies a separate path for state updates. The practical test will be whether those additions preserve the responsiveness implied by the 100-millisecond block design once real applications use them.

The first operational checkpoint is the upgrade planned roughly four weeks after Nov. 9; until then, builders must plan around the disclosed single-sequencer setup, the 5.5-day withdrawal challenge and the available read interface without the fast-write channel.