I've already written quite a lot about the new generation of performance-focused chains. Still, there's at least one more that's worthy of a write-up: MegaETH.

Given they've just launched their public testnet, letting us all experience the speed and potential of their design, it's probably a good time to take a look at what makes MegaETH special.

Before we get to that, a quick note to say there probably won't be another newsletter for a few weeks as I have a lot going on right now and likely won't have as much time to focus on this as I'd like. I'm also hoping to spend a bit of time reflecting on how I want this to evolve and fit in with my other (slightly neglected) channel Topic Crypto, as well as how I can focus on creating and growing an audience for both.

I really enjoyed this episode of Bankless covering native and based rollups.

This podcast delving into Celestia was also great.

And Bell Curve featured an interesting discussion on crypto governance.

MegaETH is yet another new chain focused on performance. In their words, they aim to be the first "real-time" blockchain offering a level of responsiveness comparable to Web2 applications. However such claims are common among these competing chains, and so it would be easy for MegaETH to be lost or overlooked in the group. Except, MegaETH is taking an approach that makes it stand out.

See, MegaETH is arguably the first project to truly lean into and exploit the full potential of a modular design.

As an L2 sat on top of Ethereum, and relying on EigenLayer's EigenDA for data availability, MegaETH can safely delegate crucial responsibilities like security and censorship resistance to those lower layers so it can focus exclusively (and aggressively) on performance.

While any L2 is free to act in this way, surprisingly few actually do it. Many of them try to retain some degree of decentralisation and they either plan on, or already do, distribute transaction sequencing and block production responsibilities across a network of nodes -- just like a traditional L1 blockchain.

However, that naturally creates a performance bottleneck and introduces latency. For a transaction or block to be properly confirmed, all of those nodes need to communicate and come to consensus: data must travel between them and that takes time.

MegaETH, in its drive to maximise performance and minimise latency, is doing away with this entirely. Their chain will rely on a single active sequencer with incredibly performant hardware to execute all transactions -- meaning no other nodes to communicate or find consensus with.

Of course, this single sequencer model raises questions about how to handle the risks and downsides of centralisation, like censorship and downtime. MegaETH addresses these in a few ways.

For instance, although only one sequencer will be active at any one time, they will regularly rotate between about 15 different sequencers run by different entities. All of these sequencers will have funds locked up on Ethereum that will be slashed if they are ever seen to misbehave or censor transactions.

In worst-case scenarios, users can rely on a force inclusion mechanisms on the L1 to ensure their MegaETH transaction is processed. Or, they can use the escape hatch to exit the chain even if the sequencers are offline.

Meanwhile, to protect against liveness failures, MegaETH will constantly have a passive sequencer ready to take over if the active one goes offline.

Altogether, these features should mitigate the worst aspects of centralisation.

At the same time, MegaETH will help to reassure users that their single sequencer is acting properly by maintaining reasonable hardware requirements for the non-sequencing nodes who monitor and verify activity on the chain, or simply keep their own, local record of the chain's state.

There are a few different types of non-sequencing nodes in MegaETH, but perhaps the most important is Provers, who actually validate the blocks published by the centralised sequencer. To keep this role relatively accessible (i.e. not dependent on the expensive, high-performance hardware used by sequencers), each prover will be tasked with a subset of the sequencer's transactions. And, they will be able to validate blocks asynchronously and out-of-order using proofs.

The additional performance unlocked by relying on a single sequencer enables MegaETH to produce blocks at an incredibly fast rate: one every 10ms. They even aim to reduce this to a block every 1 ms.

This is dramatically faster than any other chain achieves today. Solana, for instance, produces blocks every 400ms, whilst Arbitrum produces them every 250ms.

If they can pull off either of their target block times, they certainly should be offering an experience comparable with traditional Web2 speeds. Applications that require low latency,

Interestingly, MegaETH will produce a second set of blocks at a slower rate of 1 per second in a parallel chain. They refer to the 10ms blocks as mini-blocks, and the 1s blocks as EVM blocks. The design is somewhat similar to shreds on Solana or flashblocks on Base.

Both sets of blocks will contain the same transactions and offer the same guarantees, they will only differ in the amount of metadata they include. Mini-blocks have much less metadata compared to EVM blocks, which helps them propagate across the network faster to aid the chain's responsiveness.

The EVM blocks really exist to ensure compatibility with the existing Ethereum/EVM ecosystem and tooling. Developers who don't want to update their code and processes to work with the reduced metadata of mini-blocks can instead rely on the larger EVM blocks. Additionally, parameters like EIP-1559 gas pricing and block size will be set using the EVM blocks rather than mini-blocks, preserving EVM equivalence.

The other noteworthy feature of MegaETH's blocks is that they will be produced elastically, based on the demand to transact. If there are no transactions, then the chain chooses to produce no blocks rather than redundant, empty blocks.

Of course, there's much more that goes into producing blocks quickly than a single sequencer, and MegaETH has made many other lower-profile optimisations to increase speed and throughput.

For instance, like many other high-speed chains such as Monad, MegaETH will use a parallelised EVM to process multiple transactions simultaneously.

They're also rewriting a common data structure used in EVM chains.

Typically, these chains rely on something called a Merkle Patricia Trie to represent and store their state. However, updating these tries can be slow and computationally intensive, creating a performance bottleneck for high-speed chains. Therefore, MegaETH aims to build a new structure that is functionally equivalent to the Merkle Patricia Trie but much more performant.

Like Monad and Berachain, MegaETH has been quite successful in building a thriving and engaged community of builders and potential users even before the chain has launched.

Their fundraising model has played a part in this. At a time when the crypto industry has been particularly critical of the 'unfair' advantage of VCs and other investors, MegaETH has tried to make early-stage investment accessible to unprofessional, everyday investors.

After a $20m seed round led by Dragonfly Capital (and featuring big names like Vitalik Buterin), MegaETH decided to raise an additional $10m from 3200 investors on the fundraising platform Echo. The terms of this second raise matched those given to the seed investors.

Next, they sold 5000 non-transferable NFTs called Fluffles for 1 ETH each, cultivating a whitelist of potential minters from their own community and other aligned groups. These NFTs will grant their holders an allocation of future tokens. (Disclosure: I was fortunate enough to mint a Fluffle myself).

Another 5000 Fluffles will be minted in the future, with MegaMafia projects determining eligible users this time around.

The MegaMafia is the chain's flagship incubator program, helping developers launch applications that are only possible on a chain with the speed and performance of MegaETH. Members include the trading venue GTE, lending protocol Avon, prediction market Sweep, and the decentralised VPN Hop Network.

As with other pre-launch chains, it's encouraging to see such a strong community around MegaETH given the early stage of the project. However, it remains to be seen whether they can sustain the community and convert them into actual users when the chain finally launches and (perhaps more importantly) tokens start trading.

As mentioned earlier, MegaETH has just launched their public testnet. You can access it here: https://testnet.megaeth.com/.

If you need tokens, you can find a faucet here: https://testnet.megaeth.com/#3.

And find projects to check out here: https://testnet.megaeth.com/#5.

There aren't too many currently live, but I'm sure that will change in the days and weeks ahead. A good place to start might be:

• Creating and trading tokens on GTE https://testnet.gte.xyz/

• Borrowing and lending on Teko https://testnet.teko.finance/.

• Whack a Fluffle on Awa: https://awe.box/