Blast to shut down its Ethereum L2, citing costs that

Blast said it will wind down its Ethereum layer-2 network after operating costs exceeded chain revenue, telling users to withdraw assets back to Ethereum mainnet. The shutdown hits as Ethereum developers scheduled the Glamsterdam upgrade for Sepolia on Oct. 6, keeping L2 operational risk and upgrade readiness in focus.

Key Takeaways

Blast said it will wind down its Ethereum layer-2 network after operating costs exceeded chain revenue, and it told users to withdraw assets to Ethereum mainnet.
Ethereum developers scheduled the Glamsterdam upgrade to activate on the Sepolia testnet on Oct. 6, and the Ethereum Foundation said Sepolia node operators must update both execution-layer and consensus-layer clients beforehand.
September 2026 hack-loss estimates clustered around $767 million, with PeckShield tallying $766.5 million across 55 major incidents and CertiK estimating $768.4 million across 97 incidents.
Veteran trader Peter Brandt put Bitcoin’s next peak in a $300,000 to $600,000 range for late 2029 and said there is a “good possibility” the cycle low is already in.

Blast Pulls the Plug on Its Ethereum L2, Tells Users to Exit to Mainnet

Blast said it is shutting down its Ethereum layer-2 (L2), a network that processes transactions off Ethereum and periodically settles back to mainnet, after the chain’s operating costs exceeded the revenue it generated. The team said it sees no “credible path” to making the network economically sustainable and instructed users to withdraw assets to Ethereum mainnet.

The message was blunt: “Unfortunately, the economics of operating the chain no longer make sense.” For traders, the immediate operational item is not narrative, it is plumbing. Assets sitting on an L2 that is winding down become a time-sensitive bridge and withdrawal problem, especially if many users try to exit at once.

Blast’s rise and fade was fast by L2 standards. The chain launched in November 2023 with native yield on Ether and stablecoins plus a points program tied to an anticipated token airdrop, and it attracted more than $2 billion in deposits before mainnet launched in February 2024.

L2 Economics Stress-Test: When Chain Revenue Can’t Cover the Bills

An “uneconomic” L2 is not just a business-model footnote. It is a stress test of what happens when sequencer operations, infrastructure, and ongoing maintenance costs are not covered by fees and other chain revenue. When that gap persists, the failure mode is predictable: incentives shrink, service quality degrades, and eventually the operator chooses an orderly exit over subsidizing the chain indefinitely.

Blast’s own numbers frame the arc. After pulling in more than $2 billion in deposits pre-mainnet, its DeFi total value locked (TVL), the dollar value of assets deposited into its DeFi apps, fell more than 98% from its June 2024 peak, according to DeFiLlama data cited in the digest. A shrinking TVL base tends to compress fee generation and liquidity depth at the same time, which makes the chain less attractive for activity that would otherwise produce revenue.

The practical risk for active L2 users is concentrated in three places. First is withdrawal throughput and user guidance: if the exit path is “withdraw to mainnet,” traders need to know whether there are deadlines, throttles, or changing instructions, because those details determine whether exits are smooth or congested. Second is application unwind: positions in DeFi apps can be liquidated or stranded if liquidity dries up faster than users can migrate. Third is bridge and cross-chain routing exposure, where a chain in wind-down can become a weak link in a multi-hop strategy.

This is also why L2 selection keeps reverting to the same trader question: does the chain have a durable revenue engine, or is it effectively running on subsidy and attention. Blast’s statement removes ambiguity. The operator is explicitly saying the chain cannot pay its own bills.

Glamsterdam Hits Sepolia on Oct. 6: Client Updates and the Upgrade Checklist

Ethereum developers scheduled the network’s next major upgrade, Glamsterdam, to activate on the Sepolia testnet on Oct. 6. Sepolia is a public Ethereum testing network where upgrades are activated before mainnet to surface bugs without risking real funds, and the Ethereum Foundation said Sepolia node operators must update both execution-layer (EL) and consensus-layer (CL) clients before the activation.

That EL and CL requirement is the operational tell. Ethereum upgrades fail in boring ways when client versions lag, and testnets are where those coordination problems show up first. For traders, the Sepolia date is not a price catalyst by itself, but it does turn Glamsterdam from a roadmap noun into a calendar event that can generate real signals: client readiness, unexpected instability, and whether the upgrade’s components behave cleanly under load.

The digest summary of Ethereum Foundation notes highlighted three Glamsterdam components that matter for market structure even before mainnet. Enshrined proposer-builder separation (PBS) moves the handoff between specialized block builders and validators into the protocol to reduce reliance on outside middleware. Block-level access lists record the accounts and storage locations used during each block. Parallel processing support aims to let clients execute work concurrently to improve throughput.

Each of those changes is ultimately about throughput and the shape of MEV plumbing, but the nearer-term question is simpler: does Sepolia activation happen cleanly, and do operators actually update in time.

The Jay Clayton AI czar. Blast shuts Milestones Ahead

Blast’s next communications matter more than the shutdown headline. Traders need process details: whether there are deadlines for withdrawing to mainnet, whether guidance changes as the wind-down progresses, and whether bridge congestion appears as users exit in size.

The Oct. 6 Sepolia fork is the other near-term checkpoint. If client-update compliance is messy or Sepolia sees instability around activation, that tends to leak into mainnet expectations, even if the mainnet timeline is not specified in the packet.

Risk tape is still being written in the background. September 2026 hack-loss estimates came in at $766.5 million across 55 major incidents from PeckShield and $768.4 million across 97 incidents from CertiK, with the month including the Bitget hack and a $320 million Liquid Network hack where “more than $270 million” was later returned. The totals are close but the incident counts are not, and the digest itself carried inconsistent Bitget loss figures ($388 million, $387.7 million, and $387.5 million), leaving room for later reconciliation.

Policy and sentiment headlines are also competing for attention. President Donald Trump named Jay Clayton to lead a new “Super Intelligence Force,” writing that it is tasked with coordinating federal efforts “to ensure that America continues to lead the World in Super Intelligence.” Clayton is described in the digest as U.S. director of national intelligence, a detail not corroborated elsewhere in the packet.

How I’m Reading This Week’s Mix: L2 Shakeouts, Upgrade Cadence, and Risk Appetite

The part that matters in the Blast shutdown is not the brand, it is the admission that chain revenue could not cover operating costs. When an L2 operator says there is no “credible path” to sustainability, the trade is operational: get assets back to mainnet on your own schedule, not the crowd’s, and assume liquidity thins as attention leaves.

The threshold that matters on the Ethereum side is whether Oct. 6 on Sepolia is boring. If node operators update EL and CL clients cleanly and the fork is stable, Glamsterdam starts to look like a normal cadence event rather than a coordination risk. Layer that over September’s roughly $767 million in hack-loss estimates and the setup reads as risk management, not hype: the market is still paying for plumbing failures, and the chains that keep users safest are the ones that make exits and upgrades uneventful in practice.

Sources

Cointelegraph