Why I Am Excited About PolkaVM

The EVM was designed in 2013. It shows.

256-bit words because that’s what Solidity needed. Stack-based because that’s simple. Gas metering retrofitted onto operations never designed for it.

PolkaVM starts fresh. Built in 2024, it reflects what we’ve learned about blockchain VMs.

Why a New VM?

The EVM works. Thousands of smart contracts run on it. Why replace it?

Performance. The EVM is slow. JIT compilation helps but can’t fix fundamental design decisions.

Security. Stack-based execution makes control flow analysis hard. Vulnerabilities persist because the code is hard to reason about.

Tooling. The EVM is a weird target. Compilers struggle with its constraints.

The Design

PolkaVM is register-based, like modern CPUs. This maps better to hardware and enables better optimization.

It uses RISC-V as its base instruction set. Not because blockchains run on RISC-V hardware, but because RISC-V is simple, well-understood, and has excellent tooling.

Gas metering is built in from the start. Every instruction has a defined cost. No surprises.

The RISC-V Connection

RISC-V is an open instruction set architecture. It’s simple, around 50 base instructions. It’s well-specified. Compilers target it well.

PolkaVM takes RISC-V and adds blockchain-specific constraints: deterministic execution, gas metering, sandboxing.

You write Rust (or C, or anything that compiles to RISC-V). It runs on PolkaVM. The toolchain is standard.

Performance

Early benchmarks show 10-100x improvements over EVM for compute-heavy operations.

Why so much faster?

Register machines map to real hardware better. JIT compilers have decades of work on RISC architectures. The instruction set is designed for efficient execution, not blockchain-specific quirks.

Migration

PolkaVM can run EVM bytecode through translation. Not perfectly (some gas costs change) but most contracts work.

This gives a migration path. Existing contracts keep running. New contracts get the performance benefits.

What It Means

Blockchain VMs have been stuck. The EVM set the standard, and compatibility locked everyone in.

PolkaVM breaks that lock. By choosing a mainstream instruction set, it rides the wave of existing tooling instead of rebuilding from scratch.

Smart contracts become more like normal programs. The weird constraints of blockchain execution remain, but the execution model is familiar.

This matters more than it sounds. Developer experience compounds. Tools improve. Performance unlocks new use cases.

The EVM was good enough for 2013. PolkaVM is what a blockchain VM should look like in 2025.