Vitalik Buterin shared a strawmap last week. It’s not code. It’s not an EIP. It’s a single sentence: Ethereum Foundation plans a new VM for privacy and scalability. Candidates: leanISA and RISC-V. The market yawned. But this is the most structurally consequential signal since Proof-of-Stake.
Context: The Liquidity of Trust Every blockchain is a trust machine. The EVM has been the engine of that trust for nearly a decade. It’s battle-tested, but it was never designed for ZK proofs or native privacy. Every L2 today — Arbitrum, Optimism, StarkNet — runs a modified EVM or a custom VM to work around its limitations. The result: a fragmented execution layer where composability leaks like a bad smart contract. Liquidity is merely trust, tokenized and flowing — and the EVM’s constraints create friction that bleeds trust.
Core: The Architecture of The Next Iteration Buterin’s strawmap targets a radical shift: replace or supplement the EVM with a RISC-V or leanISA-based virtual machine. Why RISC-V? It’s an open-standard instruction set architecture, already used in hardware (chip design, IOT). In a blockchain VM, it enables three things the EVM cannot deliver natively:
- ZK friendliness — RISC-V’s arithmetic can be expressed in circuits far more efficiently than EVM opcodes. This means L1-level privacy without expensive trusted enclaves.
- Deterministic optimization — A lean ISA (instruction set architecture) reduces the attack surface for formal verification. Fewer opcodes = fewer bugs. This is the path to a provably secure execution environment.
- Potential hardware acceleration — If the same ISA runs on chips, validators can offload execution to dedicated RISC-V cores, increasing throughput without sacrificing decentralization.
The catch? Backward incompatibility. Every existing Solidity contract — billions of dollars in locked liquidity — would need to be recompiled, re-audited, and redeployed. This is not a weekend upgrade.
But here’s the nuance: the strawmap does not say “replace EVM”. It says “new VM”. The most likely path is a dual-VM architecture — the current EVM for legacy contracts, a new VM for next-gen dApps (privacy, ZK-heavy, high-throughput). This mirrors StarkNet’s approach (Cairo VM for native, EVM compatibility via Kakarot). But StarkNet is an L2. At L1 scale, a dual-VM introduces new attack vectors: state divergence, equivocation, and liquidity fragmentation.
Contrarian: The Decoupling Thesis The market assumes “new VM = better VM”. I see a different risk: execution debt.
In 2022, I watched Terra’s algorithmic stablecoin collapse because the team optimised for growth before verifying the tether. Here, the Ethereum Foundation is proposing to rebuild the engine while the plane is flying. The most dangerous debt is the kind no one sees — in this case, technical debt that compounds with every delay. If the new VM takes 5 years to deliver, during which the community is distracted, Solana and newer L1s (Sui, Aptos) could eat Ethereum’s liquidity share. The liquidity map is not static; it follows the path of least friction.
Furthermore, RISC-V as a VM is unproven at scale. It’s a hardware ISA, not designed for stateful smart contracts. Adapting it will require a colossal rewrite of the Solidity compiler, the Geth client, and every L2 that intends to settle in the new environment. The probability of a successful migration within 3 years is low. The probability of a hard fork over incompatibility is non-trivial.
Takeaway: Positioning for a Long Gamma Play Don’t trade this. The signal is too early. But if you’re a macro allocator, start mapping the signals: watch for a formal research paper (likely within 6 months), a GitHub repo under the ethereum org, and a discussion in AllCoreDevs. If the proposal gains consensus, accumulate ETH when narratives heat up — because a successful Lean Ethereum would solidify Ethereum’s role as the global settlement layer for another decade. If it stalls, the liquidity will flow to L1s that already deliver on execution. Structure precedes value; chaos destroys both.