The future of the internet: Inside the race for Web3’s infrastructure

People interact with open-source applications like MetaMask, Web3 games, the metaverse and DeFi protocols every day but don’t often stop to think about what happens in the background for it all to work. If we think of Web3 as a burgeoning new city, node infrastructure providers are the underlying power grid that makes operations possible.

All DApps need to communicate with blockchains, and full nodes serve billions of requests from DApps to read and write data to chains every day. We need a huge node infrastructure to keep up with vastly expanding DApp ecosystems and serve all of the requests. However, running nodes is very time and capital intensive, so DApp builders turn to providers for remote access to nodes. There is a massive monetary incentive for infrastructure providers to power as many of these Web3 ecosystems as possible, but who is winning this race so far?

The centralization problem

The fastest way to provide reliable infrastructure to power DApp ecosystems is for centralized companies to set up a fleet of blockchain nodes, commonly housed in Amazon Web Services (AWS) data centers, and allow developers to access it from anywhere for a subscription. That is exactly what a few players in the space did, but it came at the price of centralization. This is a major issue for the Web3 economy, as it leaves the ecosystem vulnerable to attacks and at the mercy of a few powerful players.

Consider that over 80% of Ethereum nodes are located in the United States and Germany, and that the three largest mining pools could come together to 51% attack the network. In many ways, today’s blockchains are a lot more centralized than we’d like them to be, in stark contrast to the ethos originally set out in Satoshi Nakamoto’s Bitcoin (BTC) white paper.

If large node providers collude, Web3 would lose all the advantages it has over Web2, from censorship-resistance to trustworthiness, and be stuck with only its disadvantages, from relatively high fees to low transactional throughput.

Not only that, but reliance on centralized providers also leaves the door open to outages. For example, an Infura outage actually forced crypto exchanges and wallets, like Coinbase Wallet, Binance and MetaMask, to suspend Ethereum and ERC-20 token withdrawals, since they couldn’t fully rely on their nodes.

It’s also worth noting that Amazon, which is the backbone of many of these centralized providers, has suffered a number of outages in the past, creating another layer of vulnerability. Ethereum’s Infura outage isn’t the only one. More recently, Ethereum’s move to Ethereum 2.0 was set back with a 7-hour outage due to the hardware failure of a single node on the network. This is a risk that truly decentralized networks don’t have to worry about.

Decentralization is a key tenet of the Web3 economy, and centralized blockchain infrastructure threatens to undermine it. For instance, Solana has suffered multiple outages due to a lack of sufficient, decentralized nodes that could handle spiking traffic. This is a common problem for blockchain protocols that are trying to scale.

And it’s not just Solana. Many of the top blockchain protocols are struggling to find a way to scale and become more decentralized. In fact, while large blockchains like Ethereum and Bitcoin have remained steadfast in the war for decentralization, smaller blockchains have lost the battle, suffering 51% attacks at the hand of overly-centralized node providers.

For instance, on June 8, 2013, Feathercoin (FTC) suffered a 51% attack. This means that a single entity was able to control more than half of the total processing power of the FTC network. This allowed them to reverse confirmed transactions and even halt new transactions from going through.

At the same time as the FTC attack, the website suffered a DDoS attack. This made it difficult for users to access information about the attack or to try and get their…