From the get-go, the goal has always been for the Ethereum network to grow into an energy-efficient, scalable Proof of Work (PoW) consensus model. This end-state has however yet to be attained, with the network thus far still undergoing various iterations in pursuit of this goal.
Whereas the current Proof of Stake (PoS) model is energy-intensive, the targeted PoW consensus mechanism is much more energy-efficient, depending on node coin holdings rather than computing power to validate on-chain transactions.
Various names were assigned to each subsequent iteration of the ETH network, including; Olympic, Frontier, Homestead, Metropolis, Serenity, etc. It is by the end of Serenity that the network is expected to handle a substantially higher number of per-second on-chain transactions while utilizing as efficient an energy model as possible.
At the end of Serenity, it is largely expected that the network would have increased substantially the number of transactions it is able to handle per second, making the Ethereum network more scalable as it grows.
1. Olympic
This first public testnet was released in early 2015, later to be replaced by the Morden upgrade, which was released with Olympic’s replacement upgrade, Frontier. The primary reason for this upgrade was that the Olympic private keys were potentially vulnerable to hacks.
2. Frontier
Named ‘Frontier’ because it was meant for early adopters and blockchain experimenters, this release was a raw as they came. It featured a bare-bones non-user friendly Command Line interface and was considered the main public network to replace Olympic. Commonly termed as Ethereum 1.0, which then pivoted to Homestead in 2016.
3. Homestead
Happened in March 2016 and is the ETH network iteration widely considered to be Ethereum 1.1, developed as a major upgrade to Frontier.
4. Ropsten
This was a Homestead-equivalent public testnet launched towards the end of 2016 to resolve issues on the older Morden network. The famous Ropsten attacks (Feb 2016) however prompted the development of the Kovan testnet upgrade, a Proof of Authority (PoA) network designed to curb such attacks in the future.
5. Metropolis
This release was supposed to be Ethereum 1.2, aimed for 2017 initially, and intended to be split into 2 phases. Primarily intended as an upgrade to the Homestead network which would open up Ethereum to decentralized applications (dApps).
6. Serenity
The Serenity upgrade will mark the beginning of the transition to Proof of Stake (PoS), a change that Vitalik opines would signify crossing past the 60% mark on Ethereum network development.
The Merge
The Serenity upgrade should be ushered in by The Merge, a long-awaited event that will make the ETH network more robust, secure, and stable while remaining maximally decentralized and settling an increased number of transactions per second.
Projected to happen midway through 2022, The Merge will be a solid step towards turning the ETH network into an L2-focused ecosystem to make it faster, more scalable, and cheaper to use. An L1 network, which Ethereum currently still is, will only get more expensive to use, as demonstrated by the high gas fees that most Ethereans have not stopped complaining about.
Although lots of development work still remains to be completed on the ETH network, The Merge is a long-planned event at which point the current Ethereum mainnet will be joined to the Beacon Chain Proof of Stake (PoS) chain. Vitalik has however revealed that a hard-fork will follow 6 months after this merge, to allow for on-chain developer validator nodes to withdraw their thus-far locked ETH.
The Surge
Following The Merge and its subsequent hard-fork, the next step on the Ethereum Grand Vision would be a Sharding Implementation, projected to split the Ethereum network into several portions (shards).
The Surge (Sharding Implementation) would be an ongoing attempt to further boost the network’s scalability and overall network infrastructure. The Surge will also improve the Ethereum network’s scalability for future roll-ups, which would have pulled the ETH network to nearly 80% full development once this Sharding Implementation has been executed.
That said, the Sharding Implementation is expected to split Ethereum network traffic into 64 new chains, spreading out the network load and taking some of the computations and storage functions off the chain. This combination of upgrades will likely increase on-chain transaction speeds while simultaneously lowering on-chain transaction costs. According to Vitalik Buterin, Ethereum co-founder and programmer, a fully-deployed Sharding Implementation should see costs going down to a maximum of 5 cents per transaction