Thesis v1.0 · published May 2026

The Valdium Thesis

Three problems are converging on the chains we use today. Each one alone would be uncomfortable. Together, they describe a system running on borrowed time. Valdium is a layer one built from genesis on the answers to all three.

Why I'm building this

I have watched the same conversation repeat itself for years. Quantum computing is somebody else's problem. Solidity is fine because the tooling around it improved. Validator hardware costs are reasonable because professionals run them. Each individual claim is defensible in isolation. Together they describe a network that is one hard breakthrough, one bad incentive, and one regulatory shift away from a very bad day.

I'm building Valdium because every chain I would want to ship a serious application on has at least one of these problems baked in at the genesis block, and you cannot patch a genesis block. You can only start over.

The chains we have will break under quantum

The problem

Every major blockchain in production today signs transactions with an elliptic-curve scheme — ECDSA on Ethereum, ECDSA on Ethereum, Schnorr on Bitcoin. All of them are mathematically demolished by Shor's algorithm running on a sufficiently large quantum computer. When that machine arrives, an attacker who can derive a private key from a public key can drain every wallet whose public key has ever touched the chain.

This is not a research curiosity. Every funded transaction reveals a public key. Every contract address derived from a key reveals one. The exposure window is not "when quantum arrives" — it's everything that has ever moved, retroactively, the moment the algorithm can run at scale.

The evidence

NISTFinalised three post-quantum standards (FIPS 203 / 204 / 205) in August 2024. The agency's own published guidance is to begin migration immediately.

IBMHas publicly committed to a 100,000-qubit machine by 2033. Internal roadmaps target cryptographically relevant scale well before that.

CRQC estimatesMost academic estimates for a cryptographically relevant quantum computer cluster between 2030 and 2040. A small but serious minority say sooner.

Harvest-nowNation-state actors are already archiving encrypted internet traffic on the assumption it will be decryptable later. The same logic applies to chain history.

What we're doing

Valdium signs every transaction with ML-DSA-65 (Dilithium3) — the NIST FIPS 204 standard for digital signatures, designed specifically to resist quantum attack. Not an optional opt-in. Not a layer two. The base protocol. From block zero.

The trade-off is real: signatures are larger (~3.3 KB instead of 64 bytes) and verification is slower. We accept both. We engineered the block format, the gossip layer, and the validator throughput targets around those numbers from day one rather than retrofitting them into a chain optimised for ECDSA.

Migration is not an option. A chain that retrofits post-quantum signatures inherits the security of its weakest historical link. Every key ever exposed is a key forever exposed. We build on PQC from genesis because that is the only honest path.

The smart-contract developer pool is starved on purpose

The problem

There are roughly 22,000 active Solidity developers and around 9,000 active Rust developers writing for blockchains. That is the pool every general-purpose chain is fishing in. Meanwhile there are over 17 million JavaScript developers on Earth, and the gap between those two numbers is the most expensive bottleneck in this industry.

The story we tell ourselves is that smart contracts need a special language — that the surface for bugs is too large, that the EVM model is too peculiar, that JavaScript is too dangerous. None of that survives close inspection. The dangerous parts of JavaScript come from the browser DOM and from network IO. Strip those, freeze the surface, and you have a clean, well-understood, deterministic language with a working ecosystem of two decades.

The evidence

Stack OverflowJavaScript has been the most-used programming language for ten consecutive surveys. Solidity does not appear in the top twenty.

GitHubJavaScript and TypeScript repositories outnumber Solidity repositories by approximately 400:1.

Audit costMedian Solidity audit pricing has held above $500/hour for three years — a direct symptom of a scarce specialist pool.

ReentrancyThe single largest class of exploits in EVM history is a footgun no language with structured async would let you fire.

What we're doing

Valdium contracts are written in plain JavaScript. The source is deployed on-chain, not bytecode — anyone can fetch a contract address and read exactly what is running. There is no compiler to trust, no decompiler to second-guess, no symbol table to lose.

Determinism by sandbox. Contracts run in a restricted V8 with no clock, no network, no filesystem, no source of randomness except what the chain provides.
Reentrancy by construction. Calls between contracts are async message-passing — there is no synchronous "call into another contract and resume" path for an attacker to recurse through.
Immutability by default. A contract has no admin key. If the deployer wants upgradability, they opt in explicitly and the upgrade path is visible to every caller before they sign.

"Decentralised" is mostly a brand claim

The problem

If running a validator requires a colocation contract, a 128 GB server, a 10 Gbps link, and a 32-coin stake, then the network is centralised — it just has a polite name for it. The number of nodes is irrelevant if all of them are inside the same five hosting providers.

The end state of every chain whose hardware floor rises is the same: a small professional cohort runs nearly all the validators, the protocol's behaviour begins to reflect their incentives, and the original promise of permissionless participation becomes marketing.

The evidence

EthereumOver 60% of mainnet validators were running on three cloud providers at peak concentration. The number has improved slightly. It has not improved structurally.

EthereumRecommended validator spec calls for 256 GB RAM, multi-NVMe storage, and a metro-fibre uplink. The realised validator set reflects exactly the hardware floor that demands.

Cosmos chainsMedian validator set size has trended down across the ecosystem for three years as professionalisation displaces hobbyists.

What we're doing

Valdium targets the median consumer laptop as the validator floor. Not a high-end gaming rig. A laptop. A Raspberry Pi 4 with a decent SSD is supported and tested.

We get there with three design choices: a small rotating BFT committee (~100 validators per epoch) so the per-block work stays bounded; two-second block times with sub-block-level work shed off the critical path; and a stake floor that prices a small operator in, not out.

The tradeoff is throughput. Valdium will not match Ethereum on raw transactions-per-second, and we don't pretend otherwise. We are optimising for the property that's expensive to recover later, not the one that's cheap to scale through a layer two.

What Valdium is, in one paragraph

A post-quantum layer one with quantum-safe signatures from genesis, smart contracts in plain on-chain JavaScript with reentrancy structurally removed, and a validator floor low enough that retail operators are a credible majority rather than a rounding error. Built on the assumption that the chains we have today are running on borrowed cryptography, borrowed language ecosystems, and borrowed claims about decentralisation — and that the right response is to start over with all three answered at the base layer.

Where we are

Public testnet — running, two-second blocks, 27 validators across three continents, 14,900+ blocks finalised under real BFT consensus.
Operator app — Electron desktop client that runs a full validator with GUI onboarding and chain sync. No terminal required.
Native wallet — Chrome extension with Dilithium3 signing, per-origin permissions, and standard transaction approval flow.
Docs and AIPs — in progress. The improvement proposal process is the same one we'll use post-mainnet.

The ask

If you build smart contracts and you've ever lost a weekend to a reentrancy bug — try Valdium. If you've wanted to run a validator and the hardware sheet said no — try Valdium. If you take quantum risk seriously — try Valdium.

The launch list is the fastest way in. Mainnet drops as a fair launch on Ethereum — no presale, no insiders, no team allocation. The mint address goes live on the official channels at the moment it does, and not one second earlier.

Next up Post-Quantum Cryptography →