Ethereum Layer 2 Benefits: Scalability, Lower Fees, Security

L2 are based on the idea that it will never make sense for every $0.82 candy bar to be broadcast to every other human on the entire planet. That every transactions by every human for everything needing to be sent to every other human and then stored forever just makes no damn sense. Simply making L1 blocks larger is just heading down that path. Could they make L1 blocks 5x bigger? Sure it would be trivial and then in a year or two when that is maxed out what then? Just making L1 larger until it runs the entire world is not scalability solution. It is hand waving the problem away.

So you either accept that or you don't. If you don't then well L2 make no sense just make the super duper god L1 and eventually everyone uses that for everything.

If you accept that the future is multiple blockchains then read on. Ethereum believes they have a very secure network so it would make sense to leverage that. This is because when you have multiple semi-independent blockchains effective security can never be higher than the weakest link.

A simpler concept than an L2 is a sidechain. You can think of sidechains as a porotype version of L2. Polygon (POS) is the best known sidechain. Polygon is its own chain. It has its own validators, its own security model, its own blocks, its own nodes. A user on a sidechain like Polygon though is now dependent on BOTH the security model for the Ethereum and Polygon. If either one fail you lose your crypto. No matter how secure you make Polygon it is still no more secure than Ethereum. Likewise even if Ethereum is bulletproof security if Polygon is compromised you still lose.

The reason for this is we can't really transfer tokens between blockchains. While it looks like we do that it is actually an illusion. If you transfer a token say 1000 USDT from Ethereum to Polygon what actually happens is actually this. First you deposit the 1000 USDT into a bridge smart contract on the the Ethereum chain. Then you take proof of that transaction to a smart contract on the Polygon chain which mints out of thin air 1000 USDT on Polygon. To go back the opposite happens. You burn (destroy) the 1000 USDT on Polygon through the smart contract on that side in a provable manner, provide that proof to the contact on Ethereum and it unlocks and transfer 1000 USDT to you.

The key thing to realize here is the USDT on Polygon only has value if there is always an equal amount locked up on the Ethereum side bridge contract. If that ever isn't the case you effectively have "counterfeit" USDT. In fact you could have your USDT stolen without it being stolen. You technically would still have the tokens they just would be worthless.

So as you can see if an attacker can compromise security on Ethereum (either the bridge smart contact itself or a cartel of malicious validators) then they can just steal the coins locked up for Polygon by taking the ones on Ethereum side and making the Polygon tokens worthless. Likewise if the Polygon side is compromised someone could fake burning tokens and trick the Ethereum side into unlocking the tokens to steal them. Now to be clear L2 also use very similar bridge contracts but tx on the L2 are written to the L1 (in compressed form will get to that later) which means the L2 can be directly validated by data on the L1.

In a sidechain security is actually the lesser of the two security models it can never be better than that. You can improve the security of both chains but it is impossible for overall security to be better than the lesser. No matter how secure Polygon is it can't be more secure than Ethereum but it could be less secure. With me so far?

L2 are an acceptance of that so they use the L1 for security. Validators on L2 are the first line or processing tx but they rollup transactions in compressed summarized form to the L1 (Ethereum). The short version is validators on the L2 are checked by the strong security model in the L1. If they cheat they can be caught on the L1. Now this does mean L2 are still not more secure than the L1 but the assumption is you would need to trust Ethereum regardless. In eventually at least in theory L2 can be no less secure than the L1.

So far we just talked about security and hopefully you accept all that on why L2 are more secure than a sidechain and so why we would prefer an L2 over a sidechain. However if writing the L2 tx on the L1 took the same space as a native L1 tx it would be pointless. L2 write a compressed summary form of tx to the L1 you can fit a lot more of them without making L1 blocks larger.

That means total tx throughput between L1 and all L2s is increased by at least two orders of magnitudes (from 10 tps to 1,000 tps). This is not hyperbole all L2 combined have already hit 100 tps simultaneously and taking individual peaks of various L2 closer to 300 tps. Pretty confident that right now with zero tweaks or optimizations we can do 300 to 500 tps if 90% of the traffic was on L2 not L1. 10 tps max on L1 alone vs 500 tps in this hybrid model. The 500 is not a hard limit, the 1,000 isn't a hard limit. With things like full blob support (not 4884 but the full danksharding coming later) potentially to 100,000+ tps. Yes at that point if needed you could increase block sizes. Larger blocks is not verboten forever but it should be the last option because it is capacity at a cost. It is the brute force option.

The future of Ethereum if you go out 10 years is likely that almost nobody transactions on it directly. That 99% of the L1 block space is staking related txs and rollups from L2s. Most exchanges and other service providers would strongly warn users against withdrawing to the L1 due to extremely high costs. If the average gas fee on the L1 is $1,000 but that is an L2 rollup which is the summarized form of 10,000 transactions so costing those individual L2 users $0.10 ea is that a problem? Gas fees on L1 10x current but tx fees on L2 just pennies.

This process can be extended to L3. Rather than be generalized L3 would probably be optimized for specific purposes or within a single entity. For example a dex not uniswap style but something like full trading desk with realtime trades and limit orders could handle 100,000+ tps related to that exchange inside an L3 and roll up only summaries taking less space than 20 native tps on the L2 which would be comparable to

I admire him for actually being honest about it.

I don't really get the point of L2s. They're supposed to help Ethereum scale, right? But at the same time the user experience is like using an entirely different network anyway. You have to set up a totally new RPC endpoint in Metamask/whatever wallet so why not just make different scalable L1s that have nothing to do with Ethereum at that point? I just don't understand what the point is of complicating things by requiring Ethereum as a base layer.

If the point is to inherit Eth's level of security by periodically settling transaction batches on it, then I still don't get why these exist. If these were separate L1's and they got popular enough for it to matter, they would still easily attain sufficient security through a large number of miners or validators.

Can anyone explain what I'm missing? They just seem like Layer 1's with extra, unnecessary steps.

The article barely covers anything, and it doesn't link to original source:

The main points:

• L2s are too slow to reach Stage 2
• L1 is already scaling sufficiently that L2s need to find meaning beyond just scaling a little faster than L1

While making a speech at the 2021 Shanghai International Blockchain Week, Ethereum co-founder Vitalki Buterin suggested that Layer 2 was the future of Ethereum scaling and the only way to safely scale the network while preserving decentralization.

Buterin revealed that the initial phases of Eth 2.0 would soon be launched, and the Eth 2.0 roadmap would offer scalability. However, base layer scalability for applications will come as the last major phase of the upgrade, which could be years from now.

Ethereum scaling is not just an idea, but a clear roadmap where many pieces of that roadmap have either been developed or already been released or are well in progress.

Efforts to make Ethereum more scalable, secure, and sustainable have been going on behind the scenes, and the Eth2 upgrades are being deployed progressively.

The first Eth 2.0 upgrade, “Beacon Chain upgrade,” was implemented in December. It introduced staking to the Ethereum ecosystem, and the Ethereum mainnet’s merge with the Beacon Chain’s proof-of-stake system will signal the end of proof-of-work consensus. The Merge is scheduled to take place in 2022.

After the Merge, “sharding” is next on the development plan. It is a multi-phase upgrade to improve Ethereum’s scalability and capacity through the use of shard chains to spread the network’s load across 64 new chains.

Going further, Buterin admitted that the need to make Ethereum more scalable is a top priority, but Ethereum’s native scaling (sharding) would take a long time. Therefore, “Rollups,” a Layer 2 solution that handles transactions outside the main Ethereum chain (layer 1) but posts the transaction data on layer 1 was launched to cause a significant increase in scalability up to a factor of 100.

Layer 2 solutions that have been launched so far include Polygon Hermez, Loopring, ZKSync, Optimism, and Arbitrum One. Buterin pointed out that the current cost to transfer ETH on Polygon Hermez is $0.25 compared to $6.84 on Ethereum.

Buterin admitted that Ethereum had embraced rollups as a scaling strategy for the near and mid-term future, and optimistic rollups, which are easier to create, are likely to be preferred in the short term. However, ZK rollups, which are way more complex but more secure, were likely to be preferred for the long term.

Zero-knowledge rollups (ZK rollups) run computation off-chain and submit a validity proof to the chain, and are much faster than optimistic rollups, which can take about a week for withdrawals.

There is still a lot of work to be done despite the upgrade in technology and applications to rollups, and Buterin urged application developers to actively work on adopting rollups as soon as it is safe enough.

Finally, Buterin added that non-financial applications could switch to rollups to help reduce fees as users were growing weary of high fees.

Other applications that could benefit from rollups are NFTs, Ethereum Name Service (ENS), light-clients, and decentralized autonomous organizations (DAOs).

If you were to look at the daily active addresses on Ethereum (layer 1) over the last few years, we have seen a steady increase but nothing face melting. Around a 2X which is great.

​

The additional ~200,000 addresses interacting daily represents an expanding community of users/devs/stakers. These people are likely price insensitive, hence their presence during the bear market.

However, this fails to tell the story of how Ethereum has scaled since the inception of Layer 2's.

It seems that Vitalik's vision of a 'roll-up centric roadmap' is coming to fruition. Here is a post back in 2020 when he first released these ideas for the future of Ethereum:

Fast forward to today and active users on L2's are beginning to overtake that on the base chain. In the past month, the number of active addresses on Polygon, Arbitrium and Optimism has increased by 85%. You can see below that the number of active addresses on these 3 L2's alone has surpassed that of the base Ethereum chain.

​

Active addresses on L1 Ethereum vs top 3 L2's

​

Some of those addresses on Polygon will include you avatar NFT loving dudes to just highlight one use case example of how these L2's are growing the scale on which Ethereum is used.

It is still relatively early days for L2's and the long-awaited arrival of Zero-Knowledge (ZK) rollups is about to be unleashed. This will likely drive this trend further. Importantly, all of this competition is driving fees on L2 to incredibly low values vs the base chain. Of course, you should be aware of the caveat that L2's may not supply the same security as the base chain. However, for most simple transactions you can see why people would choose L2 from the current pricing:

Overall, the scalability that L2's have brought to Ethereum is fantastic, amplifying the network effects. However, they are taking some activity away from the base chain which reduces the fee revenue for Ethereum and thus may negatively affect its valuation. How this plays out over the next few years will be very interesting. As long as enough high-value transactions remain on the base chain, this will probably not be an issue. That is something that isn't really talked about a lot so I'd love to hear your thoughts.

What impresses me about ETH, is that it's being used for the majority of decentralized financial transactions. This became clearer to me after minting my free users NFT avatar and learning the process of selling it requires gas fees to transfer the NFT. I need a balance of ETH in my wallet at all times as the currency. I had the realization of just how high volume and high demand these types of markets are operating in. They're cooler than I expected!

With that said, to keep the growth trajectory while maintaining security and fees - an endless number of second layer chains can attempt to evolve with Ethereum for a better user experience and ecosystem. It seems there is a natural competitive environment for L2s that lead to a survival of the fittest, best solution wins. Does that mean ETH can continue to improve forever? Is this reason to believe supposed "ETH Killers" are not truly a threat?

As usual with crypto news the headline is sensationalist trash.

If anyone cares about what Vitalik actually said, you can read his post here: np.users.com//comments/1quv8if/on_l2s_and_ethereum/

You have to set up a totally new RPC endpoint in Metamask/whatever wallet so why not just make different scalable L1s that have nothing to do with Ethereum at that point?

This is true because todays implementation is shit. But people can make wallets that bridge and switch networks without the user having to do anything.

You can't just say these layers can reach the same security as ETH, the Trilemma remains an unsolved problem if you look at single layers. High scalability always results in lower security and/or decentralization. Building high scalable networks on top of a highly secure and decentralized one is the best solution we have found so far.

If these were separate L1's and they got popular enough for it to matter, they would still easily attain sufficient security through a large number of miners or validators.

Rollups have a couple of big advantages here.

Firstly and most obviously, alternative L1s have to pay for their own security, which they do through issuance to validators (or miners) - - inflating away the value accrual of their asset. The amount being paid out is the security budget, and it is in effect the protocol paying for people to stake and therefore secure the network from attackers.

For a rollup, you don't need to bootstrap your own set of validators for security, you can get that from Ethereum and all you have to pay in return is a portion of the gas fees collected by users. There is no way that an Alt-L1 can compete with this proposition in term of value for money security.

Secondly, by splitting up the Ethereum ecosystem into different layers, each part only requires low powered hardware to run. You can run an Ethereum node (or validator) on a cheap (~$400) computer with a regular internet connection. You can also run nodes for the major L2s with the same type of requirements -

Contrast this with an alternative chain that is trying to do everything on L1, such as Solana. Solana requires a machine with 256GB of RAM (4x more than the PC I use for VR gaming) and a 1 GB/s up & down internet connection - which is basically unavailable however much you are willing to spend almost anywhere on Earth. Not being able to run a node means that you cannot interact with the chain directly, whether to post transactions or even to check your balance. If you just have to trust a random 3rd party and can't verify anything yourself then you've given up the one advantage crypto has over the traditional financial system - you may as well just stick to banks and Paypal.

Without any doubt. ETH would be unusable without L2s. You had to pay triple digit gas fees for a $10 transaction during the bull run, lmao.

I think essential is an understatement

Layer 2 solutions are becoming more important and more used everyday. There's many layer 2 blockchains active right now like Arbitrum, Optimism, Base, Loopring and many more.

These are important because they solve an important part of the Ethereum network, the high fees and its speed. Transactions in a Layer 2 are fast and cheap and there's a reason for that.

Layer 2 solutions are a different blockchain from Ethereum but they are also an extension of Ethereum. We can call these layer 2, rollups and this name describes exactly what is happening.

When an user does a transaction in a layer 2 blockchain, the transaction gets registered only in this seperate blockchain, but when enough transactions are done, the layer 2 will take hundreds of transactions and execute them as one in the Ethereum layer 1.

What happens is that the fee paid in the Ethereum layer 1 will be divided by everyone that made a transaction in this rollup, this way the fees paid by the users will be many times less than they are on Ethereum.

The security of this transactions are the same as the ones done on Ethereum, because reversing a transaction on layer 2 means that the same has to happen on layer 1, even if the security of the layer 2 solution is poor, Ethereum layer 1 will secure the transactions in an irreversible way.

That's why layer 2's are an extension of ETH, despite not being the same blockchain, eventually the transaction will be written in the ETH blockchain.

Can someone explain to me like I'm 5 what a Layer 2 in crypto actually is? I keep hearing about things like Arbitrum, Optimism, and the Lightning Network, but I'm lost. Which coins or blockchains even have layers? Are there only Layer 1 and Layer 2 or are there more? How do these layers work and what do they actually do for regular users? And is there any other tech in crypto that works kind of like this whole "layer" concept?

L2s are cheap and fast , u need them especially when ETH is congested and one transaction costs 20+$, on l2s it will cost pennies

Is it feasible to build on layer 1 and completely get rid of layer 2?

I see things like MATIC and Enjin's Jumpnet and other DeFi projects building their own Layer 2 solutions: is this taking away from the total value locked in the Ethereum ecosystem or are these Layer 2s still technically on Ethereum if their Layer 1 is? For instantance, ETH is valuable (in part) since you need it in order to pay gas fees and eventually will be (is?) burned on transactions. If all my transactions are happening on MATIC or Enjin's Jumpnet, not Ethereum blockchain, will this take value away from Ethereum or am I missing something since I believe these coins are still built on Ethereum. Not saying Ethereum will go to zero value, but is it correct to say that as more and more projects begin to work on their own blockchains / Layer 2s, that this will decrease the demand for ETH?

Layer 2 will handle the excess demand that l1 can't handle. It will actually be good for ethereum as it will allow for more contracts and interactions with them strengthening the network effect

We all love a good Layer 2 products, they offer interesting solutions to existing ETH problems. You could argue that they are not actually necessary if Ethereum's mainnet evolves enough to meet its scalability problems and high fees during "rush hours" which is probably the most important aspect for most users.

1. Inherent security: ETH's layer 1 is already highly secure due to decentralised nature. By relying on Layer 1 improvements, we can make sure that all transactions benefit from this high level of security without relying on potentially less secure Layer 2 solutions.
2. Simple user experience: Using only the base layer simplifies the user experience. Layer 2 project require users to understand and manage additional protocols, bridges, shuttles etc. which can be very complex and super user-unfriendly. I personally shit bricks every single time I have to bridge something. I only made one mistake once and lost $50 by bridging stuff incorrectly and I was super angry about it. Luckily it was on a testnet and I didn't lose a penny, but man I would be pissed if this happened to me on a live service. A highly scalable Layer 1 solution with low fees would make interactions with ETH ecosystem much easier for users to engage with the Ethereum network. Why do I have to have like 10 different networks on my bloody Metamask? It's just a messy user experience.
3. One big liquidity pool for everyone: Staying on Layer 1 makes it that all assets and decentralised apps operate within a single, unified ecosystem. People don't need to bridge assets across multiple layers, they don't fragment the market further and they don't dilute the liquidity pools across multiple layers. From the liquidity perspective it just makes more sense to keep it all in once place.
4. It reduces complexity for developers: Developers can focus on building their applications directly on Ethereum without needing to account for Layer 2 solutions. This reduces development complexity, developers can better use their resources, and accelerate innovation instead of wasting time building multiple layers of compatibility.
5. Sustainable growth: If Ethereum's base layer can handle more transactions efficiently, it can support a growing number of users and applications without relying on external solutions. It's more sustainable for the long-term growth by continuously improving the core protocol. Instead we have people jumping from one protocol to another.

What will be the new "flavour of the month" solution that people will flock to this time around? Scroll? Linea? Base? There are just too many projects out there. They are just cannibalising the userbase from each other, it makes no sense. Even existing project are already struggling.

What are your thoughts on this?

Volume rising on L2 is due to insufficient capacity on Ethereum. The two very will rise together.

L2 still require access to the Ethereum network. People will still need to bridge assets from Ethereum to L2 and back (both require some eth gas). People may need to move to the L1 to transfer between two different L2s. People with very large asset values (thinks billions) may prefer the higher security of Ethereum over the L2 networks.

So it is dubious L2 will be so popular than Ethereum declines. Their fates are intertwined.

For anyone looking at layer 2. I highly recommend checking out Celer Network (CELR), too!