Ethereum's Scaling Bet: Can Layer Two Hold Until the Merge?

Layer-two networks are no longer a theoretical fix — applications are already scaling on them, and fast. The open question is durability: how long will these rollups and sidechains stay relevant once Ethereum's own mainnet upgrades catch up to them?

Before answering that, it helps to define terms. "L2" describes a family of solutions built to scale Ethereum applications by moving transaction processing off the main chain (L1). Rival, more centralized blockchains can already offer quicker, cheaper transactions than Ethereum — but Ethereum's constraint is that it has to scale without sacrificing decentralization or security, which is precisely the gap L2 projects are trying to fill.

Several teams are competing to solve this, though not all are growing at the same pace. Polygon — the project formerly known as Matic — has pulled ahead in recent weeks by acting as an aggregator rather than betting on a single approach. Where a project like Synthetix relies specifically on Optimistic rollups, Polygon's strategy is to fold in multiple scaling techniques at once. Major DeFi platforms including Aave, Curve, and Stake DAO have already integrated Polygon to reduce transaction costs and wait times for their users.

Mechanically, users lock Ethereum-based tokens into a Polygon bridge contract, transact freely within the Polygon sidechain, and later withdraw back to Ethereum's main chain — retaining Ethereum-level security while sidestepping the network's congestion-driven fees.

It might seem intuitive that a fully upgraded Ethereum 2.0 would make L2 solutions unnecessary. Polygon founder Sandeep Nailwal pushed back on that idea in a Coindesk interview, arguing that Eth2's arrival won't sideline his network. His case: Ethereum 2.0 is designed around 64 shards, each functioning roughly like Ethereum does today. If proof-of-stake alone lifts today's single chain from its current 13 transactions per second to around 50 tps, multiplying that across 64 shards yields roughly 3,200 tps. Nailwal's follow-up question was whether 3,200 tps is genuinely sufficient if Ethereum is meant to become the base settlement layer for global finance — his own answer being that Eth2 will be roughly 64 times more scalable than Ethereum is now, while demand could be a thousand times larger than current levels, meaning L2 scaling stays necessary regardless.

Not everyone agrees with that framing. It is possible 3,200 tps proves adequate for long enough that L2 usage fades on its own — and if demand eventually exceeds that ceiling, the response might simply be an "Ethereum 3.0" rather than a permanent reliance on L2. Other commentators expect a longer-term division of labor instead: L1 handling base-layer issuance and settlement of protocol assets, L2 handling microtransactions and providing censorship resistance. Under that model, core assets and contracts would live on L1, while L2 layers handle token distribution and related, analogous systems.

Whichever way the L1/L2 relationship ultimately settles, the pivotal event is unquestionably Ethereum 2.0's rollout — often referred to as "the merge." That process began on December 1, 2020, with the launch of the Beacon Chain, which introduced proof-of-stake to Ethereum and laid the groundwork for the next phase: shard chains. Ethereum.org notes that researchers are actively working to speed up the merge timeline, and that it will likely arrive sooner than originally projected. The same source explains that shards will expand Ethereum's capacity for storing and accessing data, but will not themselves execute code — with the implementation details of that separation still being worked out. That kind of open-ended language is a useful reminder of just how early-stage this technology remains.

There's a useful parallel in how people once talked about the early internet — many assumed it would simply become another media channel, coexisting alongside television and radio. Instead, it absorbed them. The expectation here is similar: crypto won't stay confined to currencies. Between L2 and Eth2, the infrastructure is being laid for what could become the largest speculative marketplace in existence — dwarfing what's already visible on BSC today.

The next wave of retail day traders and full-time yield farmers could bring uncomfortable attention from tax authorities, and some protocols stand to grow their market capitalization to previously unthinkable levels. Cheap, fast, and reliable transactions would also open the door to genuine everyday and microtransaction use cases for crypto, reinforcing the growth of this parallel financial system.

Ethereum isn't competing in a vacuum, either — well-capitalized, technically sophisticated rivals like Solana, Avalanche, and Elrond are all racing toward the same scaling goals, and Ethereum's first-mover advantage, while significant in an industry that moves this fast, does not guarantee it the win. It's more likely that no single chain "wins" outright; different networks will specialize in speed, security, privacy, or decentralization respectively, producing a competitive landscape where users pick the tool that fits their needs.

Looking back from where things stand now, the progress made under the weight of high gas fees and limited throughput has been considerable, and as those constraints ease and crypto becomes a normal part of daily life for more people, it's difficult not to be optimistic about what's coming next. As the Hopium Diaries put it: the drive toward financial freedom isn't going anywhere, and the only real uncertainty is the path to mass adoption, not whether it happens. Transaction costs need to keep falling and throughput needs to keep rising — and once that shift is complete, the window to get ahead of it will have already closed.