Ethereum returns to L1 narrative, Solana Consensus accelerates

The ultimate dream of Ethereum is to be a world computer, coincidentally, Alpenglow is too.

Written by: Zuo Ye

Ethereum has begun supply-side reform.

After the dream of the infinite garden shattered, Vitalik restrained the social debt of L2/Rollup and took a more proactive stance to defend the L1 track. The “speeding up and reducing costs” plan for the Ethereum mainnet has been put on the agenda. The shift to Risc-V is just the beginning, and how to catch up with or even surpass Solana in terms of efficiency will become the main focus moving forward.

Solana continues to expand its consumption demand scenarios.

Solana’s answer is Scale or Die, resolutely sticking to the path of strengthening L1. In addition to the Firedancer developed by Jump Trading entering the deployment process, at this year’s Solana conference in New York, the Anza team’s Alpenglow consensus protocol took the lead, attracting everyone’s attention.

The ultimate dream of Ethereum is a world computer, coincidentally, Alpenglow is too.

20% Security Consensus in the Era of Large-Scale Nodes

Starting with Bitcoin, the number of nodes and their distribution are synonymous with measuring the degree of decentralization of a blockchain network. To avoid centralization, the security threshold is set at 33%, meaning no single entity should exceed this proportion.

With the boost of capital efficiency, Bitcoin mining ultimately moves towards mining pool clusters, and Ethereum becomes the main stage for Lido and CEX. Of course, this does not mean that mining pools and Lido can control network operations; under the model of “maintaining the network - earning incentives/management fees,” they have no malicious motives.

Image description: The scaling law of blockchain nodes, image source: @zuoyeweb3

However, measuring the health of a network must consider its scale. For example, in a small group of 3 people, a 2/3 majority is required to be considered effectively operating. Pursuing merely the minimum security assurance of 1/3 is meaningless; the remaining two can easily collude, making the cost of wrongdoing very low while the benefits of wrongdoing are very high.

If it is a large-scale network of 10000, as Etherscan shows the current scale of Ethereum nodes, there is no need to pursue a 2/3 majority vote. Beyond the incentive model, most nodes are not familiar with each other, and the coordination cost of collusion between Lido and Binance is too high.

If we decrease the number of nodes and the consensus ratio, can we “speed up and reduce costs”?

Everyone should be able to think that Alpenglow thinks so too and is ready to do so, maintaining the scale of 1500 nodes for Solana while reducing the security consensus to 20%. This not only improves the confirmation speed of nodes and allows nodes to earn more mainnet incentives but also encourages the expansion of node scale, such as expanding to around 10,000.

It could either be the effect of 1+1>2, or it could breach the existing security mechanisms, both of which are possible.

However, I really like it. Solana should follow the American chain, conspiracy group, and centralized route, participating in the public chain competition as the opposite of Ethereum.

Modified Turbine, or moving towards DPoS

The theoretical idea of Alpenglow is that in the era of large-scale nodes, a strong consensus is not needed because in a PoS mechanism, malicious actors need to mobilize an enormous amount of capital to take control. Even at a scale of 20%, based on current prices, Ethereum would require 20 billion dollars, and Solana would also require 10 billion dollars.

With $10 billion, it’s not good to do something, it’s crazy to control the blockchain, and it’s going to be countered by the remaining 80% of nodes, unless it’s a state action.

In practice, Alpenglow roughly divides the entire process into three parts: Rotor, Votor, and Repair (for a more detailed division and workflow, refer to their paper). To some extent, Alpenglow is a deep transformation of the Turbine mechanism, so we will briefly introduce Turbine.

Image description: Block broadcasting mechanism, image source: @zuoyeweb3

Turbine is the block propagation mechanism of Solana, simply put, Turbine is the messenger that disseminates block information to achieve consensus confirmation across all nodes.

In the early design of Ethereum, block broadcasting was done through the Gossip protocol, which is named in Chinese as “谣言、小道消息”. Any node could act as the initial messenger, and P2P communication could occur between any nodes, ultimately achieving the effect of global confirmation, akin to “I have a message, I only tell you, and you must not tell anyone else,” and eventually, the whole world will know.

We temporarily skip Turbine and compare it with Ripple’s DPOS mechanism. Ethereum/Bitcoin have achieved the ultimate in peer-to-peer communication; what is the cost of that?

It is extremely slow and expensive, which is quite understandable. The aimless and random pairing between nodes, while being more robust and more resistant to censorship, will take longer, and the expense comes from the consideration of node costs. Since the dissemination of information among nodes is slow and they need to maintain their own operation, this cost will naturally be passed on to users in the form of Gas Fee.

Extreme decentralization will inevitably lead to expensive and slow accompanying ailments.

Ripple’s DPOS has gone to the other extreme, initially setting only 21 supernodes across the entire network, an extreme centralization completely solves any side effects, and now the mainnet nodes have reached 179.

Turbine, on the other hand, moves towards a middle state, neither adopting Ethereum’s Gossip mechanism nor going down the wrong path of reducing the number of nodes, but instead implementing a hierarchical propagation of network nodes:

1. In each cycle, nodes are categorized into Leader, Relay, and more ordinary nodes, only the Leader node can send block broadcast information.

2. A small number of Relay nodes continue to broadcast the information to more ordinary nodes after receiving it, and this process is called Turbine Tree, resembling a tree with clear main and secondary branches.

In Alpenglow, the variant of the protocol is called Rotor, which essentially involves the orderly dissemination of block messages, with any Leader or Relay nodes being non-fixed.

Votor is a node confirmation mechanism, for example, in Alpenglow’s vision, the first round of node voting reaches 80%, meeting the minimum of more than 20%, it can be directly and quickly passed, if the first round of voting is higher than 60% and lower than 80%, the second round of voting can be opened, and the final confirmation can be exceeded again if it exceeds 60%.

If it doesn’t work, then go through the Repair mechanism, but I personally feel that this is similar to the challenge period of the Optimistic Rollup, and if it really comes to this point, there is a high probability that the agreement will be over, and the FDIC will not be able to stop the run on Silicon Valley Bank.

Unlike the violent stacking of hardware and software resources to increase bandwidth, Alpenglow’s starting point is to reduce the block consensus generation process.

If the data block is as small as possible, such as limited to about 1500 bytes at present, the generation time is short enough, for example, in the current test, the limit can run to 100ms, which is 1% of the current 10s. (1 s = 1000 ms, but this is doubtful, it is really practical on a large scale, the conditions will not be so ideal, but it is also terrifying).

Conclusion

After MegaETH, the existing L2 basic volume reaches its endpoint. With no Solana support for SVM L2, the Solana mainnet has a real demand for continued expansion. Only with mainnet TPS overwhelming all competitors can the concept of Solana as the Ethereum killer be fully realized.

Alpenglow is not limited to Solana; theoretically, any PoS chain, including Ethereum, can utilize its mechanisms. Similar to what we introduced earlier with Optimum, existing blockchain research has reached a technical edge and urgently requires more assistance from computer science and even sociological concepts.

IBM once claimed that the world would only need five mainframes in the future. If the internet built on HTTP-TCP/IP is one, BTC is one, and Ethereum is one, then there is not much space left for Solana.