3 min read

Networks that Implement a Heavy VM at L1 will always be Plagued by Scalability Issues

Networks that Implement a Heavy VM at L1 will always be Plagued by Scalability Issues

(Any views expressed in the below are the personal views of the author and should not form the basis for making investment decisions, nor be construed as a recommendation or advice to engage in investment transactions.)

The following statements were made by Mr. Kaine in Telegram on December 20, 2022 about Zenon Network.

The whitepaper points out two sources of PoW: one CPU bound (ASIC resistant) and the other one ASIC friendly. The key is balancing them. The design principles of NoM's architecture are very important in the long term development of the network.

A simple and robust L1 with minimal features will always outperform over-engineered and complex designs that try to accomplish too many things at once. The consensus is the most important part of a distributed network because it distributes trust in a set of untrusted peers that need to coordinate to achieve a common objective: maintaining the state of the ledger.

There are many flavors for different variants of consensus protocols and usually we want add to the mix Sybil-resistant algorithms such as PoS and PoW. PoW is the most pure form of Sybil-resistance. It was designed as anti-Sybil mechanism back in the '90s. PoS is a weaker anti-Sybil mechanism than PoW, but coupled with a classical consensus protocol makes it a very good candidate to Bitcoin's PoW.

But pure PoS based consensus protocols are flawed from an economic/societal perspective that undermine decentralization and in some cases, certain security thresholds. PoW evolved into two opposite flavors: ASIC-friendly and ASIC-resistant.

Bitcoin is ASIC-friendly because its sole purpose is to convert energy as efficient as possible. But ASIC-resistant algos are important to keep participation as decentralized as possible. 1 CPU = 1 vote was just an assumption of Satoshi. Hal clearly envisioned a widespread Bitcoin adoption at industrial scale.

So we have BFT consensus based on virtual voting (outlined in a paper in '93), PoS, (dual) PoW and now let's talk about the ledger structure. The idea behind a dual ledger system is to decouple consensus from chain weight. This is very important because consensus and chain weight are fundamental for a L1.

Bitcoin has consensus coupled with chain weight: "longest chain of most accumulated proof of work". NoM has consensus decoupled from chain weight (added when users are performing tx with PoW). None of top L1s are designed to take into account all those variables and are aiming to "solve" the scalability trilemma.

In other networks, users pay a fee to make transactions and usually miners are rewarded. In NoM, users pay with Plasma (that usually involves PoW) to make transactions and by doing so, they add weight to the ledger, thus effectively securing it.

The dual-coin economy also takes into account "greedy staking" and other "the rich getting richer" schemes and provides long-term sustainability for active network participants.

Bitcoin does exactly this: it is indeed minimal (non-Turing complete) and very robust. Networks that implemented a heavy VM (EVM/WASM) at L1 will always be plagued by scalability issues in the long run.

There are a lot of things to do, but rest assured that you are building on a solid foundation with a long term vision.

During some of the Q&A Mr. Kaine made a few statements about ZK rollups and Unikernels that are no longer available to me in Telegram:

  • ZK-proof rollups are part of the strategy to keep the L1 minimal and robust. Unikernels can be implemented on top, as well as other ZK rollup solutions that will support Turing-complete smart contracts.
  • But obviously we can do even better with the implementation outlined in the whitepaper. Coupled with 10-100x more using ZK rollups.