Solana’s dominant automated market maker (AMM) Raydium hit back Monday on rumors that major volume driver Pump.Fun was preparing to launch its own AMM.

Abandoning Raydium whole hog would be a «strategic miscalculation» for the massively popular — and profitable — memecoin factory, core contributor InfraRAY said in a post on X. He cast doubt on the notion that Pump.Fun could replicate its success if it swaps Raydium out for in-house trading infrastructure.

Token investors dumped RAY en-masse this weekend after hawkeyed observers noticed Pump.Fun was apparently testing its own AMM, presumably with the intent to replace Raydium’s longstanding liquidity pools as its platform of choice. Such a move would shake up the economics of decentralized token trading on Solana.

Right now, Raydium, the chain’s largest AMM platform, captures trading fees generated by Pump.Fun memecoins that «graduated» from the launchpad to its own pools. The arrangement — in place since Pump.Fun’s earliest days — has been a financial boon for Raydium

But it also leaves Pump.Fun out of the long-term upside of the tokens its users create. That’s not to say it’s making nothing: Pump.Fun has amassed half a billion dollars on the fees it collects from early-stage token launches, one of crypto’s grandest warchest.

Raydium is currently generating over $1 million in fees every day from trading across all its liquidity pools, not just those of Pump.fun tokens. That said, over 30% of Raydium’s daily trading volume comes from Pump.fun tokens, according to a Dune dashboard, meaning a good share of its fees could dry up if Pump.Fun switches away.

«100%, revenue hit is real,» InfraRAY said in a message to CoinDesk. But he cautioned that the market’s 30% haircut on RAY tokens was «overblown» and partially due to SOL’s own weakness.

He said any pivot to a new AMM could hit myriad issues: inadequate supporting infrastructure, low demand for migrated tokens, a flop on volume at launch.

«I think that’s a real risk they are overlooking but I could be wrong,» InfraRAY said.

Pump.Fun co-founder Alon Cohen declined to comment.

U.S. authorities have seized about $31 million in crypto tied to the 2021 hack of Uranium Finance, according to a Monday X post from the Southern District of New York (SDNY).

According to the post, the seizure was the result of a joint effort between SDNY and Homeland Security Investigations (HSI) in San Diego. A spokesperson for SDNY did not return CoinDesk’s request for comment before press time, and no further details about the seizure or any related investigation were immediately available.

Uranium Finance was essentially a clone of automated market maker (AMM) Uniswap deployed on Binance’s BNB chain (then called Binance Smart Chain). In April 2021, a hacker exploited a bug in Uranium’s pair contracts to steal $50 million in various tokens. At the time of the incident, the Uranium Finance hack was one of the largest monetary exploits in decentralized finance (DeFi) history.

Read more: Binance Chain DeFi Exchange Uranium Finance Loses $50M in Exploit

After the exploit, the hacker attempted to launder a portion of the funds in a variety of ways, including using crypto mixer Tornado Cash, depositing small amounts of crypto into centralized exchanges, and, according to blockchain sleuth ZachXBT, perhaps through purchasing rare and highly valuable Magic: The Gathering trading cards.

Uranium Finance shuttered after the hack, leaving victims without answers or financial restitution. The partial recovery, which comes nearly four years after the initial attack, offers the first glimmer of hope for victims to see some of their money returned.

Ethereum’s Pectra upgrade went live on the Holesky testnet on Monday but failed to finalize in the expected time.

Pectra was activated on the Holesky testnet at 21:55 UTC (4:55 p.m. ET), but did not initially finalize according to blockchain data.

Finality is the state in which, once a transaction is confirmed and added to a block, it is immutable and cannot be reversed. A testnet is a network that copies a main blockchain (in this case Ethereum), and is used to test upgrades or new code before it goes to the main network.

It is not immediately clear why the Pectra upgrade did not finalize on Holesky. Ethereum developers were discussing Monday over the Eth R&D Discord channel what the issue could be.

This is not the first time an upgrade has not finalized on an Etheruem test network. In January 2024, when the developers were testing the Dencun upgrade, the hard fork did not initially finalize on the Goerli testnet.

What is Pectra?

The Pectra hard fork combines together 11 major upgrades, or «Ethereum improvement proposals» (EIPs), into one package. At the heart of this is EIP-7702, which is supposed to improve the user-experience of crypto wallets. The proposal, which was scribbled by Ethereum co-founder Vitalik Buterin in just 22 minutes, will allow wallets to have some smart contract capabilities, as part of a broader strategy to bring account abstraction to Ethereum — a concept that makes the usability of wallets a lot less clunky.

Another key proposal, EIP-7251, will allow validators to increase the maximum amount they can stake from 32 to 2,048 ETH. The proposal is supposed to ease some of the technicalities that validators who stake ETH face today: Those that stake more than their 32 ETH have to spread that across multiple validators, making the process a bit of a nuisance. By lifting the maximum stake limit and combining those validators, it could speed up the process of setting up new nodes.

Holesky is the first of two testnets to run through a simulation of Pectra. The next test is supposed to occur on the Sepolia testnet on Mar. 5. But according to Christine Kim, a Vice President of Research at Galaxy, developers could delay it depending on the scale of today’s issue.

After Pectra goes live on both testnets, developers will ink in a final date to activate the upgrade on mainnet.

Pectra was originally on track to be Ethereum’s biggest upgrade to date, and it’s the first big change to the blockchain in almost a year. Developers decided that Pectra was too ambitious, and they agreed to split the original package into two.

Read more: Ethereum Developers Finally Schedule ‘Pectra’ Upgrade