Vitalik Buterin has called for a dramatic rethinking of Ethereum’s architecture. In a series of posts and proposals throughout 2025, he outlined a plan to simplify Ethereum’s execution and consensus layers, proposing a replacement of the Ethereum Virtual Machine (EVM) with a RISC-V–based virtual machine. At the same time, he emphasized a need to “make Ethereum’s base layer as simple as Bitcoin’s,” expressing a desire for long-term protocol resilience, maintainability, and provability. This move is not merely cosmetic. Ethereum’s complexity, once its strength, is increasingly viewed as a liability in an era where performance, security, and scalability must coexist with verifiability. As Ethereum seeks to simplify, Bitcoin is moving in the opposite direction—using zero-knowledge (ZK) cryptography to grow in functionality without compromising its minimalist base. Meanwhile, Cardano continues to develop as a hybrid approach: deterministic, UTXO-based like Bitcoin, yet programmable and expressive like Ethereum. These three chains—once philosophically distinct—are now converging on a shared realization: complexity must be carefully managed. Whether at the base layer or in layered designs, the future of blockchain infrastructure will be shaped by how effectively each platform balances functionality and simplicity. Ethereum’s Complexity Headaches Ethereum’s flexibility comes from the EVM, a Turing-complete stack-based virtual machine built around 256-bit operations and a broad set of opcodes. This architecture, while versatile, has become increasingly difficult to maintain, optimize, and scale. It introduces security challenges (such as reentrancy vulnerabilities), makes gas estimation tricky, and complicates scaling solutions like ZK-rollups that must prove EVM execution correctness. Vitalik highlighted this in a 2025 post, stating, “Over the years, we have slowly added complexity… a 256-bit virtual machine that over-optimized for specific cryptographic use cases and added precompiles for features barely being used.” He further noted that “even a motivated high schooler can understand Bitcoin’s core rules… Ethereum has reached the point where only a very small group of experts can fully grasp the system.” The EVM’s inefficiency is especially problematic for zero-knowledge applications. Rollups and zkEVMs simulate EVM execution in circuits, but proving even simple transactions can be orders of magnitude slower due to the need to mimic the EVM’s layered complexity. Ethereum’s protocol also lacks inherent mechanisms for ordering fairness, leading to widespread issues with MEV (Maximal Extractable Value), where validators can reorder or censor transactions for profit. And although efforts like Proposer-Builder Separation and encrypted mempools are underway, these are complex mitigations layered on top of an already intricate base. A New Execution Model: RISC-V and Simplification Vitalik’s proposal centers on removing accumulated complexity by replacing the EVM with a new virtual machine, potentially based on the open-source RISC-V architecture. Unlike the EVM, RISC-V is a standard instruction set used in real-world hardware and widely supported by programming toolchains like C, Rust, and LLVM. “The RISC-V spec is absurdly simple compared to the EVM,” Buterin noted. He argued that such a move could eliminate the need for much of the current EVM infrastructure, including precompiles, specialized opcodes, and bytecode interpreters. Instead, developers could write contracts in standard languages, and compilers could produce code that runs efficiently and is easier to prove in ZK circuits. Current zkEVMs already translate Ethereum operations into RISC-V internally before generating proofs. Rebuilding Ethereum around RISC-V would reduce prover overhead dramatically, with Buterin estimating “100x+ performance improvement in many cases.”. It would also make formal verification and multi-language support easier, reducing barriers for developers and client teams alike. Transitioning away from the EVM is a long-term goal. In the interim, Ethereum may adopt a dual VM system, running both the EVM and the new VM in parallel. Eventually, the EVM could be deprecated, perhaps reimplemented as a smart contract on the new VM. Buterin’s broader goal is to pare down the protocol into something simpler, more auditable, and sustainable. “Instead of many small 1.5x improvements, do one big change and get a 100x gain,” he wrote, comparing the effort to replace EVM with the drawn-out process of deprecating opcodes like SELFDESTRUCT. On the consensus side, similar efforts are underway. Ethereum’s developers are exploring “3-slot finality,” an alternative to the current epoch-based mechanism. This proposal would collapse consensus logic into a faster, more comprehensible model, improving both efficiency and simplicity without sacrificing security. Bitcoin’s Expansion Through Zero-Knowledge Proofs While Ethereum simplifies to scale, Bitcoin is scaling through cryptography, not complexity. Its base-layer protocol remains deliberately simple, based on unspent transaction outputs (UTXOs), basic scripts, and Proof-of-Work. But around that minimal core, developers are building increasingly capable tools, especially using zero-knowledge proofs. BitSNARK, developed under the BitcoinOS project, enables zk-SNARK verification directly on Bitcoin without any change to its consensus rules. This allows DeFi logic, asset issuance, and smart contract execution to be conducted off-chain, with the final proof of correctness posted to Bitcoin as a standard transaction. Bitcoin advocate Edan Yago called this a breakthrough, saying it enables “a complete smart contract platform secured by Bitcoin’s base layer.”. With BitSNARK, a Bitcoin transaction can verify that a computation, such as a swap, a loan, or a vote, was executed correctly, without requiring Bitcoin to process every step. What makes this remarkable is that it doesn’t require a fork. Existing opcodes are used to verify compact ZK proofs, preserving Bitcoin’s hard-won consensus and neutrality. Developers can now build trustless applications like DEXs, atomic swaps, and even rollups entirely on top of Bitcoin. Other ZK initiatives, like BitVM and ZeroSync, follow similar goals, creating ways to execute arbitrary logic or sync full node data using succinct proofs. While some of these still require soft forks or opcode extensions, BitSNARK is already live, marking a significant shift in Bitcoin’s capabilities. Thus, while Bitcoin’s base remains static, its ecosystem is expanding. Through ZK rollups and off-chain computation, Bitcoin is becoming more programmable, arguably more like Ethereum, while Ethereum attempts to become more like Bitcoin in its simplicity and reliability. Cardano: Balancing Structure and Flexibility Cardano stands in between. Architected as a UTXO-based blockchain with Turing-complete scripting, it combines Bitcoin’s deterministic accounting model with Ethereum-like programmability. Its eUTXO model attaches scripts and datums to transaction outputs, enabling secure and predictable contract execution. Unlike Ethereum, Cardano avoids a shared global state. Transactions operate on local data and produce deterministic results, which helps eliminate classes of vulnerabilities like reentrancy. A transaction in Cardano either passes all validation checks or fails. There’s no partial execution or side effects. Cardano’s scripting language, Plutus, is based on Haskell and emphasizes formal verification. While powerful, it’s known for being difficult to use. To address this, the community introduced Aiken, a modern smart contract language with cleaner syntax and better tooling, which still compiles to Plutus Core. Cardano’s layered design and deterministic execution make it resilient, but it was initially criticized for “concurrency issues”. The difficulty of many users interacting with the same contract. In practice, developers have built workarounds using off-chain logic and transaction batching. Cardano’s model excels in parallel transaction processing, but it requires a different mental model from the account-based design of Ethereum. Now, Cardano is integrating zero-knowledge cryptography through a project called Starstream, which adds a zkVM that operates alongside its base ledger. Starstream is a WASM-based virtual machine where contract logic is executed off-chain, and proofs of correctness are posted on-chain. This design resembles Ethereum’s ZK-rollups but is deeply integrated with Cardano’s UTXO structure. “Starstream doesn’t abandon the UTXO model — it empowers it,” one developer commented. Every contract interaction can produce a composable ZK proof, offering scalability and privacy without complicating the base protocol. By adopting ZK and building on WASM, Starstream shares philosophical common ground with Ethereum’s RISC-V proposal: both prioritize simple, well-supported VMs that are ZK-friendly. However, Cardano’s solution is layered rather than a base protocol overhaul. Converging Technologies, Diverging Strategies Ethereum, Bitcoin, and Cardano are all responding to the same pressures—scalability, security, usability - but through very different strategies. Ethereum is restructuring its core, eliminating legacy architecture and replacing it with something more sustainable and provable. Its goal is to retain flexibility while reducing risk and inefficiency. Bitcoin is enhancing functionality externally. Through ZK rollups and protocols like BitSNARK, it can support smart contracts and DeFi without changing its core logic. It prioritizes absolute stability and decentralization. Cardano blends both philosophies. Its eUTXO model enforces deterministic logic and clean execution, while it adds programmability and scalability through layered design, new languages, and ZK tools like Starstream. Interestingly, all three chains are adopting zero-knowledge cryptography to solve scalability and privacy. Ethereum is refactoring itself to make ZK native. Bitcoin is embracing it externally. Cardano is embedding it into a modular architecture. The differences lie not in what features are adopted, but where and how complexity is handled. Vitalik Buterin summarized Ethereum’s motivation clearly: “The EVM is growing in complexity, and much of that complexity has proven unnecessary. We should focus on a simpler protocol with fewer moving parts. Make the base layer lean. Prove correctness with ZK. Keep the future open.” As these chains evolve, simplicity is emerging not as a constraint but as a design advantage. The coming years may show that the best-performing blockchains aren’t the ones with the most features, but the ones with the most deliberate architecture.
