A little under four years ago we announced Cloudflare's first experiments in web3 with our gateway to the InterPlanetary File System (IPFS). Three years ago we announced our experimental Ethereum Gateway. At Cloudflare, we often take experimental bets on the future of the Internet to help new technologies gain maturity and stability and for us to gain deep understanding of them.

Four years after our initial experiments in web3, it’s time to launch our next series of experiments to help advance the state of the art. These experiments are focused on finding new ways to help solve the scale and environmental challenges that face blockchain technologies today. Over the last two years there has been a rapid increase in the use of the underlying technologies that power web3. This growth has been a catalyst for a generation of startups; and yet it has also had negative externalities.

At Cloudflare, we are committed to helping to build a better Internet. That commitment balances technology and impact. The impact of certain older blockchain technologies on the environment has been challenging. The Proof of Work (PoW) consensus systems that secure many blockchains were instrumental in the bootstrapping of the web3 ecosystem. However, these systems do not scale well with the usage rates we see today. Proof of Work networks rely on complex cryptographic functions to create consensus and prevent attacks. Every node on the network is in a race to find a solution to this cryptographic problem. This cryptographic function is called a hash function. A hash function takes a number in x, and gives a number out y. If you know x, it's easy to compute y. However, given y, it's prohibitively costly to find a matching x. Miners have to find an x that would output a y with certain properties, such as 10 leading zero. Even with advanced chips, this is costly and heavily based on chance.

While that process is incredibly secure because of the vast amount of hashing that occurs, Proof of Work networks are wasteful. This waste is driven by the fact that Proof of Work consensus mechanisms are electricity-intensive. In a PoW ecosystem, energy consumption scales directly with usage. So, for example, when web3 took off in 2020, the Bitcoin network started to consume the same amount of electricity as Switzerland (according to Cambridge Bitcoin Electricity Consumption Index).

This inefficiency in resource use is by design to make it difficult for any single actor to threaten the security of the blockchain, and for the majority of the last decade, while use was low, this inefficiency was an acceptable trade off. The other major trade off accepted with the security of Proof of Work has been the number of transactions the network can handle per second. For instance, the Ethereum network can currently process about 30 transactions per second. This ultra-low transaction rate was acceptable when the technology was in development, but does not scale to meet the current demand.

As recently as two years ago, web3 was an up-and-coming ecosystem with minimal traffic compared to the majority of the traffic on the Internet. The last two years changed the web3 use landscape. Traffic to the Bitcoin and the Ethereum networks has exploded. Technologies like Smart Contracts that power DeFi, NFTs, and DAOs have started to become commonplace in the developer community (and in the Twitter lexicon). The next generation of blockchains like Flow, Avalanche, Solana, as well as Polygon are being developed and adopted by major companies like Meta.

In order to balance our commitment to the environment and to help build a better Internet, it is clear to us that we should launch new experiments to help find a path forward. A path that balances the clear need to drastically reduce the energy consumption of web3 technologies and the capability to scale the web3 networks by orders of magnitude if the need arises to support increased user adoption over the next five years. How do we do that? We believe that the next generation of consensus protocols is likely to be based on Proof of Stake and Proof of Spacetime consensus mechanisms.

Today, we are excited to announce the start of our experiments for the next generation of web3 networks that are embracing Proof of Stake; beginning with Ethereum. Over the next few months, Cloudflare will launch, and fully stake, Ethereum validator nodes on the Cloudflare global network as the community approaches its transition from Proof of Work to Proof of Stake with “The Merge.” These nodes will serve as a testing ground for research on energy efficiency, consistency management, and network speed.

There is a lot in that paragraph so let’s break it down! The short version though is that Cloudflare is going to participate in the research and development of the core infrastructure that helps keep Ethereum secure, fast, as well as energy efficient for everyone.

Proof of Stake is a next-generation consensus protocol to secure blockchains. Unlike Proof of Work that relies on miners racing each other with increasingly complex cryptography to mine a block, Proof of Stake secures new transactions to the network through self-interest. Validator's nodes (people who verify new blocks for the chain) are required to put a significant asset up as collateral in a smart contract to prove that they will act in good faith. For instance, for Ethereum that is 32 ETH. Validator nodes that follow the network's rules earn rewards; validators that violate the rules will have portions of their stake taken away. Anyone can operate a validator node as long as they meet the stake requirement. This is key. Proof of Stake networks require lots and lots of validators nodes to validate and attest to new transactions. The more participants there are in the network, the harder it is for bad actors to launch a 51% attack to compromise the security of the blockchain.

To add new blocks to the Ethereum chain, once it shifts to Proof of Stake, validators are chosen at random to create new blocks (validate). Once they have a new block ready it is submitted to a committee of 128 other validators who act as attestors. They verify that the transactions included in the block are legitimate and that the block is not malicious. If the block is found to be valid, it is this attestation that is added to the blockchain. Critically, the validation and attestation process is not computationally complex. This reduction in complexity leads to significant energy efficiency improvements.

The energy required to operate a Proof of Stake validator node is magnitudes less than a Proof of Work miner. Early estimates from the Ethereum Foundation estimate that the entire Ethereum network could use as little as 2.6 megawatts of power. Put another way, Ethereum will use 99.5% less energy post merge than today.

We are going to support the development and deployment of Ethereum by running Proof of Stake validator nodes on our network. For the Ethereum ecosystem, running validator nodes on our network allows us to offer even more geographic decentralization in places like EMEA, LATAM, and APJC while also adding infrastructure decentralization to the network. This helps keep the network secure, outage resistant, and closer to the goal of global decentralization for everyone. For us, It’s a commitment to helping research and experiment with scaling the next generation of blockchain networks that could be a key part of the story of the Internet. Running blockchain technology at scale is incredibly difficult, and Proof of Stake systems have their own unique requirements that will take time to understand and improve. Part of this experimentation though is a commitment. As part of our experiments Cloudflare has not and will not run our own Proof of Work infrastructure on our network.

Finally, what is “The Merge”? The Merge is the planned combination of the Ethereum Mainnet (the chain you interact with today when you make an Eth transaction) and the Ethereum Beacon Chain. The Beacon chain is the Proof of Stake blockchain running in parallel with the Ethereum Mainnet. The Merge is much more significant than a consensus update.

Consensus updates have happened multiple times already on Ethereum. This could be because of a vulnerability, or the need to introduce new features. The Merge is different in a way that it cannot be made by a normal upgrade. Previous forks have been enabled at a specific block number. This means that if you mine enough blocks, the new consensus rules apply automatically. With the Merge, there should only be one state root for the fork to happen. The merge state root would be the start of the Ethereum Mainnet PoS journey. It is chosen and set at a certain difficulty, instead of block height, to avoid merging a high block with low difficulty.

Once The Merge occurs later this year in 2022, Ethereum will be a full Proof of Stake ecosystem that no longer relies on Proof of Work.

This is just the start of our commitment to help build the next generation of web3 networks. We are excited to work with our partners across the cryptography, web3, and infrastructure communities to help create the next generation of blockchain ecosystems that are environmentally sustainable, secure, and blazing fast.