Scaling Ethereum — is the future rollup-centric or multi-chain?

An in-depth consideration of Ethereum’s position as the dominant smart contract network (and how its scaling woes may threaten it) 

Beyond metaverse mania and NFT madness, one of 2021’s dominant themes has been the so-called Layer-1 wars. Where once Ethereum enjoyed unparalleled network effects that positioned it as the dead cert in the smart contract platform race, its inability to scale fast enough to meet user demand for block space has increasingly brought its dominance into question. 

As new users flock to the cryptocurrency industry in search of the gains of those who came before them, a new breed of blockchain has absorbed those alienated by Ethereum’s prohibitive network fees. Yes, Solana, Avalanche and other Layer-1s are less decentralized than the smart-contract first-mover. However, users and developers seem to increasingly value utility over ideology. 

In this OKX Insights in-depth article, we revisit Ethereum’s gas problem — and how rising transaction fees have forced newcomers to explore opportunities beyond the original smart contract platform. We consider Ethereum’s slow progress toward scaling alongside the voracious growth of alternate Layer-1s, ultimately asking if Ethereum’s network effects are strong enough to attract new users and developers in the face of rapidly growing competition. 

Ethereum’s rising gas bill

Earlier this year, OKX Insights published an in-depth analysis of Ethereum’s fee structure. In it, we argued that high network fees serve as a bottleneck for future mainstream adoption. Since then, rising prices across crypto assets and exotic blockchain use cases have brought new activity to the dominant smart-contract platform — creating yet greater demand for the network’s finite block space. Already under pressure from an expanding decentralized finance sector, the summer’s speculative mania surrounding nonfungible tokens served to highlight the issue more clearly than ever.  

One might assume this growth in transaction costs is a sign of rising network adoption — and, in part, it is. However, looking at other Ethereum usage metrics tells a rather different story. 

For example, daily active address statistics from BitInfoCharts.com show only modest growth since the beginning of the year. In early 2021, there were just under 700,000 unique daily active Ethereum addresses. As of mid-December, there are around 860,000 active addresses — an underwhelming 23% increase, particularly given the hype surrounding all things cryptocurrency this year. 

Indeed, comparing the above graph to that below, we see an almost exact correlation between the gas price peak and the high point for daily active wallets on the network. Both occurred during the local market top in May, as the price of ETH surpassed $4,000 for the first time. 

Meanwhile, statistics for the much-more-recently-launched Avalanche and Solana networks show extremely aggressive growth. Figures for the Avalanche C-Chain show daily active addresses reaching a high of 10,590 during the early months of 2021. As interest in alternate Layer-1s grew throughout the year, that figure has regularly exceeded 100,000. 

Accurate daily active address metrics for Solana are difficult to come by. However, we can infer similar growth from other sources. For example, a recent report by digital currency asset manager Grayscale compared the growth of the popular Solana wallet Phantom to that of Ethereum’s Metamask. It found that it took Metamask from May 2019 to October 2020 to grow from 300,000 to 1.3 million monthly active users. Meanwhile, Phantom wallet’s monthly active users grew from 200,000 to 1.2 million between August and October 2021.  

Crypto newcomers compromise on decentralization for utility and profit

Periods of rising cryptocurrency prices have historically been those in which adoption surges. Understandably, many new users are drawn to the industry by the speculative nature of the emerging asset class. 

Since speculation rather than ideology seems to attract users to the industry, it is fair to assume that many crypto newcomers in search of potentially life-changing gains are willing to compromise on decentralization. Those who adopted crypto during the 2017 bull market were prepared to accept similar decentralization tradeoffs to speculate on Ethereum and its initial coin offerings in greater numbers than those who had adopted Bitcoin in previous years. 

Similarly, 2021’s newcomers were again willing to speculate on more-centralized platforms that also presented potentially greater upside. Not having the luxury of previous cycle gains to pay the frankly extortionate Ethereum transaction fees, they instead adopted those networks offering a user experience more in line with their own economic circumstances. Those networks willing to compromise on decentralization to provide a more welcoming user experience have benefited greatly over the last 12 months. 

Ethereum block space is expensive precisely because the network strives to maximize decentralization. Meanwhile, block space on Solana, Avalanche and other Layer-1 blockchain networks is cheaper simply because the computer systems required to store the blockchain itself are more powerful. 

Although there is little serious debate over the relative decentralization of Bitcoin, Ethereum and newer networks like Solana and Avalanche, a revealing experiment earlier this year conducted by long-time Bitcoin proponent and altcoin critic Jameson Lopp highlights the point well. Lopp attempted to run a node on several popular blockchain networks using a high-end consumer computer system. Having struggled to sync an Ethereum node with his six-core, 32 GB machine, the same task on newer chains optimized for speed and low transaction fees — such as Solana — often proved impossible. In concluding, Lopp stated:   

“It’s pretty clear that some of the newer ‘high performance’ networks have almost none of their users running nodes and apparently few folks are even looking into the requirements for fully validating the system. This is concerning because an important aspect of any ‘decentralized’ network’s security model is if you can operate as a sovereign peer on the network who needs not ask permission to use the protocol.”

As Lopp alludes, new users likely haven’t spent hours pouring over resources extolling the virtues of decentralization. They simply want to make the kind of percentage gains their predecessors enjoyed — who were in many cases originally drawn in as speculators themselves. Apparently, if that means compromising on decentralization, then so be it. 

While the likes of Solana and Avalanche have optimized for greater user capacity on the base blockchain at the cost of decentralization, there is some early evidence that these networks will eventually be plagued by those issues that today appear to be hindering Ethereum adoption. In Avalanche’s case, C-Chain fees rose to levels that saw some users compare the network to Ethereum during periods of heightened usage. Meanwhile, Solana has yet to experience such a fee spike but did suffer around 17 hours of downtime in September 2021 as bot-generated transactions flooded the network.    

Where does this leave Ethereum?  

Concern over future scaling is nothing new for Ethereum developers. Indeed, the network’s cofounder Vitalik Buterin has been discussing ways to scale the leading smart-contract network without compromising on decentralization since its 2014 inception. Until recently, the solution was a multi-year upgrade known as Serenity, or Ethereum 2.0, which was detailed by Buterin at Ethereum Dev Con 2018.

Although considerable progress has been made, Ethereum users have now patiently waited for the network’s ongoing upgrade to roll out for several years. With the last stages of the multipart upgrade still reportedly a couple of years away and competition from alternate Layer-1s growing, the narrative has recently switched to one dominated by a family of scaling solutions known as rollups.

Rollups’ origins can be traced back to Buterin’s earliest musings on how to scale Ethereum. However, the scaling solution took center stage in October 2020 when the network’s cofounder posted “A rollup-centric Ethereum roadmap” to the Ethereum Magicians developer forum. Reasoning that “base-layer scalability for applications” is still “years away,” he made the case that “the Ethereum ecosystem is likely to be all-in on rollups” in the short- to mid-term.  

Rollups are a form of Layer-2 scaling technology where transaction execution occurs away from the main blockchain. Simply put, computations are performed off-chain, batched together and proofs of their validity are posted on-chain. Since less on-chain computation is required, gas expenditure is much lower than possible when executing transactions on Layer-1.  

There are two flavors of rollups in use on Ethereum today — Optimistic Rollups and zkRollups. In the former’s case, transactions occurring off-chain are presumed valid until proven otherwise. While this makes them comparatively simple to implement, they suffer from a couple of drawbacks:

• Firstly, there needs to be a reasonably long window in which network participants can challenge and prove fraudulent transactions. This means withdrawing funds from an Optimistic Rollup cannot be instantaneous, harming the overall user experience.
• Secondly, Optimistic Rollups also require at least one honest node to be online to perform any computations challenged as fraudulent. While incentive structures do encourage nodes to perform such validity proofs, strictly speaking, there are no absolute guarantees that there will be a node available when required.    

Meanwhile, zkRollups involve much more advanced cryptography. As such, they are a more unstable technology and porting existing EVM-compatible code to them is a much greater challenge — for now, at least. Despite believing that further development will reduce potential attack vectors, the greater risk associated with zkRollups prompted Buterin to comment in January 2021: 

“I personally would much rather trust $10 million of my own money to an EVM Optimistic Rollup than to an EVM zkRollup for at least the next couple of years, but in the long-term, zkRollups I think are going to be everything.”

Although implementations of Optimistic Rollups and zkRollups have only gone live on Ethereum’s mainnet during 2021, the technology has experienced reasonable adoption already. Arbitrum, the most successful Optimistic Rollup to date, launched in late May. By November, it had reached almost $3 billion in total value locked. Meanwhile, the dYdX decentralized exchange launched as a zkRollup in February. By October, it had exceeded $1 billion in TVL. 

Enter the “modular blockchain”

Ethereum’s pivot from its focus on Ethereum 2.0 to its rollup-centric roadmap has prompted even greater theorizing about the future of blockchain technology. Indeed, the narrative espoused by several of the industry’s thought leaders is a focus on scaling via departmentalization.  

Simply put, nodes on layers above the base blockchain specializing in different tasks should enable greater efficiency without compromising the decentralization that makes a blockchain innovative. In one of many articles detailing Ethereum’s scaling roadmap, pseudonymous writer on the subject and independent researcher Polynya explained:  

“Departmentalization is a proven concept since the dawn of the industrial revolution — why try to do everything yourself, when collaborating with specialized entities could yield greater efficiency? I believe the same is true of blockchains.” 

Today, Layer-1 blockchains attempt to handle execution, data availability, and security on a single layer. Each node must store the entire blockchain history, be powerful enough to execute computations and ensure that the chain can securely come to consensus. Polynya calls these networks “monolithic blockchains,” reasoning that such an approach makes current implementations “cripplingly inefficient.” Just as the eighteenth-century Scottish economist Adam Smith argued in An Inquiry into the Nature and Causes of the Wealth of Nations, dividing labor between specialist artisans resulted in greater output, Polynya believes the same holds true for blockchains. 

For Polynya, Ethereum makes the ideal base layer for what they term a “modular blockchain” implementation. In correspondence with OKX Insights, they explained:  

“Bitcoin and Ethereum are the only two highly secure, decentralized, proven and credibly neutral networks, backed by sound money. Since Bitcoin cannot yet support rollups, that leaves Ethereum as the best settlement layer for rollups.” 

Similarly, Joseph Lubin, the cofounder of the Ethereum Foundation, believes the network’s future is one where it serves as a central hub to which various Layer-2 scaling solutions settle. Speaking to Decrypt at the Miami DCentral conference in November, he stated:   

“Ethereum is going to be the blockchain of blockchains. It’s going to be the major digital asset settlement layer, it’s going to be the coordination layer for many different Layer-2 technologies.”

Essentially, the narrative surrounding scaling Ethereum has shifted — at least, in the medium term — from one focused on making base layer computation affordable for the masses to one in which humans are no longer the network’s primary users. Indeed, transaction fees are an important part of a blockchain’s security model and rock bottom fees mean transaction validators are poorly incentivized to stay honest — an issue that alternative Layer-1s may eventually come up against. Presuming rollups and other future Layer-2 scaling methods are adopted, they will compete with one another for block space on the base blockchain, settling batches of transactions executed on other layers. The users of these scaling solutions will essentially share the mainchain fees required to benefit from Ethereum’s security. 

While Polynya believes almost all activity will eventually take place on these Layer-2s, they concede that there will still be a class of users willing to pay a premium for the admittedly greater security the base blockchain guarantees. They explained to OKX Insights: 

“It’s likely the canonical Ethereum execution layer will always have some activity from those that want maximum security and to whom transaction fees are totally negligible — whales, governments, institutions. […] I expect 99% of people to transact directly on rollups, or eventually, through smart wallets and interfaces that abstract it further for seamless UX.”

Too late to rollup?

Had recent progress on rollups and other Ethereum scaling solutions, such as Polygon, arrived earlier, it is entirely possible that Solana, Avalanche, Cardano and others would have struggled to attract users. However, with decentralized application and smart contract development demonstrating blockchain use cases that appeal to increasingly wider audiences, demand for cheap block space simply not available on Ethereum has grown — forcing users to look elsewhere.

While Ethereum enjoyed unparalleled network effects throughout 2018 to 2020 bear market years, 2021’s influx of new users has seen development activity shift to the non-EVM compatible Solana, in particular. Although concrete metrics for developer activity across different blockchains do not exist, some sources do suggest significant growth of those building on the alternate Layer-1. 

In a recent tweet, Natasha Che, a Ph.D. in macroeconomics and occasional writer on blockchain technology, highlighted the rapidly growing Solana developer community. The data, sourced from analytics provider Santiment, only shows contributors to different blockchain’s GitHub repositories and does not account for those building on-chain applications or developing Ethereum scaling solutions. However, it does highlight massive growth in the number of people working on Solana — particularly since mid-2021, when Ethereum’s gas prices reached their peak. 

Ethereum still reigns supreme by most metrics — but that dominance appears to be waning. Statistics from multichain decentralized finance analytics platform DeFi Llama show that Ethereum accounts for around 62.5% of TVL in DeFi applications. At 366, it boasts by far the most expansive DApp ecosystem, too. When added to the 140 apps on Polygon, 49 on Arbitrum and 14 on Optimism, the number of applications on Ethereum and its scaling solutions look even more impressive. 

However, the same source shows significant TVL growth on other chains. On Jan. 1, 2021, Ethereum accounted for almost 97% of funds deposited to DeFi applications. As of Dec. 22, 2021, it accounts for less than two-thirds of DeFi’s TVL — with BSC, Terra, Solana and Avalanche now accounting for 7.1%, 5.2%, 4.9% and 4.7%, respectively. 

Helping non-Ethereum networks expand — alongside the increasing diversity of applications on alternate Layer-1s and Ethereum’s gas issues — is the growing bridging ecosystem between different Layer-1s. As more users become familiar with bridges, the moat around each network shrinks, making a multichain future increasingly likely. A September 2021 article by Dmitriy Berenzon, a research partner at 1kxnetwork, highlighted the growth of cross-chain bridges. In it, he listed more than 40 different projects currently working on bridging solutions, before going on to stress the importance of such protocols to the wider crypto ecosystem’s development.    

While Ethereum strives to become the settlement layer of a diverse rollup ecosystem, it is too early to tell if new users will turn to alternate networks or to its Layer-2 scaling solutions. Maximum decentralization is a burden in many ways. It results in expensive block space that prices all but the wealthiest users out, as well as many prospective applications. However, in the face of a potential future state-level attack, the ability for users to verify transactions themselves is crucial. 

We have yet to see such an attack and, in its absence, an entirely centralized database is a far more efficient solution than any blockchain can be with the technology currently available. However, presuming that there will never be a concerted effort to disrupt blockchain networks is a dangerous road to go down — particularly given their disruptive potential. As such, Polynya is adamant that there is no other option but to focus on energies on building the most decentralized blockchain ecosystem possible, as challenging as it may be: 

“This is complex, bleeding edge tech that’ll take time to mature. It’ll be a bumpy transition to a modular, multi-layered world, but it’s the only way the blockchain industry can attain global scale. We don’t have a choice!”

Not a part of the OKX community yet? Sign up to get started.