Important points
- MegaETH leverages Ethereum as its premier blockchain execution environment.
- MegaETH stress tests achieved 55,000 transactions per second.
- Layer 2 solutions that replicate Layer 1 services face security challenges.
- Ethereum’s strengths are critical to the entire smart contract ecosystem.
- Ethereum’s scaling strategy is moving from layer 2 to layer 1 enhancements.
- Recent changes in Ethereum’s roadmap represent a logical shift.
- Regulatory pressure is pushing some rollups toward centralization.
- Ethereum may return to higher transaction costs as activity returns to layer 1.
- The current low transaction costs are due to the movement of activities to Layer 2.
- The energy spent on Layer 2 development is essential to the evolution of Ethereum.
- Layer 2 evolved from sharding and allowed for diversity and experimentation.
- A zk-based optimistic anti-fraud system ensures transaction accuracy.
Guest introduction
Lei Yang is the co-founder and CTO of MegaLabs, the team building MegaETH, a high-performance layer 2 of Ethereum with sub-10ms block times and 100,000 transactions per second. He earned his PhD in computer science from MIT in 2024 for his research on blockchain consensus and networking that enabled end-to-end systems with 80,000 transactions per second in 2018-2019.
Why MegaETH chooses Ethereum
Mega uses Ethereum because it provides the most performant blockchain execution environment.
— Ray Yang
The reason Mega uses Ethereum is not because it’s meh as you know, but because it allows them to build the most performant blockchain possible.
— Ray Yang
- MegaETH achieved 55,000 transactions per second during mainnet stress testing.
We ran a stress test on mainnet, so everyone could test running 55,000 transactions per second.
— Ray Yang
- The stress test demonstrated MegaETH’s ability to handle large volumes of transactions.
We looked at it and it was really amazing…it trains extremely fast, has incredible compute, and can handle crazy levels of activity.
— Ray Yang
- MegaETH’s performance metrics are highly reliable as the mainnet environment closely emulates real-world launch conditions.
Flip the switch and you’re running mainnet. It’s like having the exact same server, same IP address, same settings, same management key.
— Ray Yang
Evolution of Ethereum’s scaling strategy
- Ethereum’s scaling strategy is moving from relying on Layer 2 to enhancing Layer 1 capabilities.
The original version of Vision no longer works. Ethereum itself needs to scale in its original version. Basically, we were offloading a lot of this to l two, but now that l one is scaling, we don’t need that anymore.
— Ray Yang
- Recent changes in Ethereum’s roadmap represent a logical pivot rather than a value pivot.
I think of this as a pivot rather than a values pivot, but I use the word pivot from the perspective of the roadmap itself, in terms of what was originally envisioned in the rollup-centric roadmap and where it is now.
— Ray Yang
- The move to more centralized solutions with some rollups is an understandable, although unfortunate, response to regulatory pressures.
In a sense, I think this is a sad reality similar to Moloch’s victory. To be honest, everyone has different incentives…In the stage 1 rollup, they said they didn’t want to be in the stage 2 point gap because they were in a regulatory environment that preferred to be more centralized.
— Ray Yang
- As activity returns to Layer 1, Ethereum will eventually return to higher transaction costs.
I think it’s a bit of an overreaction to say Ethereum doesn’t need another 2…we’ll soon be back to a world where File 1 has $200 transactions.
— Ray Yang
Role of Layer 2 Solutions
- Building a Layer 2 that replicates a Layer 1 service is technically uninteresting and introduces security issues.
Trying to build and replicate the exact primitives provided by Layer 1 is pretty uninteresting… It’s actually very difficult to build a secure Layer 2, and if all you want for Layer 2 is to adopt Ethereum’s security out of the box, I think you’re better off building Layer 1.
— Ray Yang
- The energy spent developing the Layer 2 roadmap is not wasted, it is essential to Ethereum’s evolution.
I really don’t think the energy we spent on this board is wasted. I think Mega is possible because Ethereum has been following this Layer 2-centric roadmap for several years.
— Ray Yang
- Layer 2 evolved from the concept of sharding and enabled diversity and experimentation in the Ethereum ecosystem.
The original idea was that Ethereum would run and operate all the shards. Why not delegate it to several teams to ensure diversity, competition and experimentation?
— Ray Yang
- Layer 2 solutions include mechanisms to ensure censorship resistance and user termination options.
First, Layer 2 does not censor users. This means that if a user’s transaction is maliciously excluded from Layer 2, they can always go back to Layer 1 and send the connection there, forcing the Layer 2 sequence to include the connection.
— Ray Yang
Economic sustainability in blockchain
- Chain fees are not a viable business model for Mega $ETH.
I don’t think it’s cool that you guys can do 11,000,000,000 transactions per second. I don’t think you guys are going to reach sustained levels of 11,000,000,000 organic transactions per second anytime soon.
— Ray Yang
- The economic sustainability of blockchain systems must prioritize user accessibility over short-term profits from fees.
I think this is actually a short-term idea because the way Mega grows is by allowing everyone to come and use my ID for as cheap as humanly possible.
— Ray Yang
- Current approaches to blockchain revenue generation need to evolve beyond simply issuing tokens.
We need to build something like some kind of actual business model, but at least we got to stablecoins before it became what it is now.
— Ray Yang
- The mechanism for monetizing stablecoins on the Mageve platform allows for reinvestment on the chain without harming end users.
What we basically said is that if you have something like this native stable coin usdm, and your application uses usdm, you get a tibo yield on that right, and that goes into Maggie Eve’s balance sheet…
— Ray Yang
The future of rollups and governance
- Stage 2 rollup requires a commitment to immutability of governance logic, which comes with significant risks.
Stage 2 is basically saying that decommissioning the Security Council code is for lifetime immutability…basically saying that one part of the design of the Layer 2 rollup will be permanently fixed for the lifetime of Ethereum, the lifetime of the universe.
— Ray Yang
- Reaching Stage 2 of the rollup may take longer than previously expected, potentially exceeding two years.
I think it is irresponsible to claim that we will reach the second stage within a year, or even within two years.
— Ray Yang
- Trusting the Security Council in the first step could lead to potential abuses of power.
We must also trust that the Security Council does not intend to abuse its power. Because I think within a year we’ll be…certainly in Stage 1, but actually, insisting on labeling something like Mega Ether as Stage 1 is probably not a good thing.
— Ray Yang
- AI could play a key role in validating the accuracy of software, which is essential to reaching the second stage.
I think what AI is best at is proving things that are difficult to create but easy to verify: software code.
— Ray Yang
Impact of AI on blockchain
- The next billion users of blockchain technology may actually be AI agents.
Another possibility, which has recently come to my attention, is that the next billion users may actually be AI agents.
— Ray Yang
- The poor user experience of cryptocurrencies for humans may actually benefit software agents.
If you believe that part of the reason we don’t adopt more cryptocurrencies is because the UX is terrible…all the UX problems we have are actually advantages for small software agents.
— Ray Yang
- We need to start prioritizing agent users in our blockchain experience.
Maybe we need to really start prioritizing agent users in some of our blockchain experiences.
— Ray Yang
- Agents have unlimited energy to try different transaction paths, unlike humans who give up after a few tries.
To support the agents in need…we need block space that is cheap enough for agents to experiment and error…agents have enormous amounts of energy…humans don’t just experiment.
— Ray Yang
The role of proximity markets in blockchain
- Ethereum’s fine-grained auctions in block order are impractical in systems with very short block spacing.
The fundamental problem for us is that the block spacing is very small. In our case, the block interval is 10ms, so running an auction at this kind of fine granularity is completely unworkable.
— Ray Yang
- We believe that the close market model is more effective than micro auctions when it comes to ordering trades.
What we’re trying to achieve is…we call it a proximity market…running these options at such intervals that a lot of people decide on it.
— Ray Yang
- Working with a sequencer allows optimized trading algorithms to operate with minimal delay.
You have seats to co-locate with your sequencer, and once that’s done, you can select the algorithms you want to run and run them on these seats, which are basically virtual machines right next to your sequencer, so the quote you get from your cloud provider is down to 1ms…
— Ray Yang
- Incentive structures should encourage high-frequency traders to be physically located near sequencers to increase market liquidity.
The reason for setting the block interval to 10 ms is to create an efficient and highly liquid market. Since the market needs to come to the sequencer, I think this is both an incentive structure and a more practical way to make efficient priority assignments when trading in real-time.
— Ray Yang
MegaETH innovation promotion strategy
- mega $ETH‘s approach of cultivating its own app ecosystem is essential to avoid redundancy between chains.
You can’t get into a situation where you have a bunch of repetitive applications that exist on every other chain. What’s the point of spending the last three years building this if there’s literally nothing new about this train?
— Ray Yang
- mega $ETH is actively seeking founders to build their own applications to strengthen our ecosystem.
All we did was basically just find founders and start convincing them to build something that was frankly cool.
— Ray Yang
- There is a huge brain drain in the crypto industry, with many potential founders leaving for AI due to poor user experience.
I think a lot of founders who could have built great apps couldn’t do it in previous cycles because their UX was rubbish, and they just let AI do it in this cycle.
— Ray Yang
- If we remain trusted and neutral and do not actively encourage development, we risk leaving new applications undeveloped.
What I’m more afraid of is that we’ll run out of developers… We can actually be incredibly neutral. If you do that, you won’t be able to run any more apps.
— Ray Yang
The future of token distribution and ownership
- The current approach to token distribution in the market is flawed and unfair.
This is an issue we saw, and we know that cryptocurrencies want to be unique, so we thought it wasn’t ideal.
— Ray Yang
- Price discovery is moving from public markets to private markets, leading to unequal access for public participants.
I think that was a pretty great thing… a lot of price discovery moved into private markets, and as a result, as you know, public participants didn’t necessarily have equal access to assets.
— Ray Yang
- Traditional ownership models in technology do not apply to cryptocurrencies, where users want to be the owners of the applications they use.
I think it’s perfectly fine that it’s not exactly the case with cryptocurrencies, where people have historically wanted to be the owners of the applications I use, but it just felt very unfair.
— Ray Yang
- The approach of allowing users to participate in ownership has proven successful, as evidenced by high participation rates in funding rounds.
We basically just doubled down on that theory…80% of Echo users tried to join…we ended up oversubscribed by about 25x.
— Ray Yang
