Second-generation blockchains such as Ethereum were marketed as faster, cheaper, and more scalable than Bitcoin. Third-generation chains like EOS and Qtum promised even greater optimizations. Yet today, these networks are grappling with the same fundamental bottleneck: competition for scarce on-chain resources leads to soaring fees and congestion. At last week's annual Ethereum developers’ conference (Devcon) in Osaka, scaling dominated the agenda. Many attendees expressed deep concerns about Ethereum’s ability to meet growing demand. Eth 2.0, the heralded upgrade from Proof-of-Work to Proof-of-Stake, remains years away, and some participants now concede it may never materialize. This deadlock has prompted rival smart contract platforms like Qtum to position themselves as more scalable alternatives.
Over on EOS, similar problems have accumulated. Unlike Ethereum, EOS is not primarily targeting decentralized finance, but it shares a common challenge: increasing demand for finite resources. On Ethereum, those resources manifest as block space, which pushes up gas prices. On EOS, they take the form of computational resources: RAM (virtual storage) and CPU (the time block producers allocate to transactions from a specific account). Surging and erratic computational costs on EOS have forced developers to seek their own scaling solutions. While the architectures of Bitcoin, EOS, and Ethereum differ substantially, the core truth remains: on-chain resources are limited, and there is an open market competing for space. Just as driving during rush hour is slower and less fuel-efficient, transactions during peak network times become slower and more expensive. Fixing this problem calls for out-of-the-box thinking.
Virtual CPUs and Off-Chain Transactions
For EOS, scaling salvation has come via Liquidapps, whose vRAM product has now been complemented by a vCPU counterpart. This approach takes precious resources off-chain to a separate network of nodes that perform computation at low cost, before broadcasting the verification back to the EOS main chain. Although currently deployed on EOS, the same technology can be applied to Ethereum or even Bitcoin Cash, offering developers a way to create decentralized applications (dapps) that require cheap storage.
For Ethereum, the most promising scaling solution is Plasma, which can handle hundreds of transactions per second and now supports smart contracts. As co-founder Jinglan Wang noted, DeFi projects claiming they don’t need scaling solutions is like a New Yorker saying they don’t need a MetroCard during rush hour. Plasma is a layer-2 solution, analogous to Bitcoin’s Lightning Network. Sidechains, state channels, and rollups are also being explored.
Dapp Developers Must Choose Wisely
Developers pondering the best network for launching dapps face tough choices:
Ethereum boasts a large ecosystem of users, developers, and enterprises, but it operates near capacity, and network fees have been rising for months. Scaling solutions like Matic (now Polygon) can help alleviate the pressure.
EOS offers free transactions at the point of access, making it more consumer-friendly. However, popular dapps risk exposing their developers to rising computational costs. Secondary solutions such as vRAM can mitigate this.
Qtum recently underwent its first hard fork, adding a new Ethereum Virtual Machine (EVM) that enhances its smart contract capabilities while retaining the UTXO model pioneered by Bitcoin. It essentially offers Ethereum-like functionality without the scaling problems—albeit with a smaller ecosystem at this stage.
Horizontal vs. Vertical Scaling
Moving transactions off-chain—whether to sidechains or layer-2 solutions—comes with compromises. Generally, it reduces decentralization and transaction finality. Advocates argue that a micropayment does not require the same security level as a $1 billion Bitcoin transaction. Gaming and gambling dapps, for instance, can comfortably use products like Liquidapps’ vRAM for off-chain storage or Plasma for low-cost transactions.
To use a traffic analogy: side streets can help bypass freeway congestion, but they are not designed for 16-wheelers. Blockchain scaling solutions are therefore better described as scaling options—choices that suit some projects but are unacceptable to others.
Vertical scaling increases throughput by raising block or node capacity. Bitcoin Cash (BCH) took this approach: simply adding more lanes to the highway. It has proven simple and effective so far. Horizontal scaling offloads as much work as possible from the main chain to third-party solutions that add extensibility.
Despite millions of dollars and tens of thousands of hours invested in blockchain scaling, we still do not know which solutions will prevail. Bigger blocks, more sidechains, building up the stack, or developing horizontally—through trial and error, a path will be found to prepare blockchains for the mass adoption that all crypto advocates see as inevitable. For now, the architects of so-called next-generation blockchains are learning a lesson bitcoiners learned long ago: on-chain, there is no such thing as infinite scalability.