1 · Introduction
The gap is measurable before it is arguable. Bitcoin sustains roughly 7 transactions per second and Ethereum's base layer ~15–30, against a mainstream payment network's ~65,000 (Visa) — while IDC estimates ~79 zettabytes of data generated by ~42 billion connected devices in 2025, exactly the kind of high-volume, low-value record that would benefit most from being verifiable. (Full figures and sources: Evidence & References.)
The usual reading of this gap is "blockchains are slow." The diagnosis here is different: the wrong data is placed at the wrong security level. A single "like", a view, one line of IoT telemetry is not worth an entire consensus network's attention — so today it stays locked inside private servers, unverifiable by anyone outside. Forcing everything onto one monolithic chain makes cheap data pay the price of expensive data; consensus-priced chains cannot absorb the data ocean, no matter how fast they get.
The approach: tier by value. Every piece of data answers one question:
Does this data need double-spend protection or composability?
That binary answer splits the system into 2 security Domains (invariant), which spread into 3 functional Layers and 5 fine-grained Tiers — the expensive cryptographic guarantees concentrated on the data that truly carries value, while the high-volume edge only needs a tamper-evident, immutable record.
Concretely, this whitepaper contributes three things most existing systems lack:
| # | Contribution | Why it is different | Specified in |
|---|---|---|---|
| 1 | A 2-Domain boundary enforced in code | The UAC primitive makes the data Domain structurally incapable of holding value, so "don't put value in the wrong place" is a guarantee, not a guideline. | Architecture · The Seven Primitives |
| 2 | A hard READ/WRITE split for cross-chain interaction | Verifying that something happened is cheap and effectively unlimited; changing state is deliberately scarce and priced accordingly. | Cross-Chain Model |
| 3 | Value-proportional security by Tier | Every asset carries a minimum security class, and assurance scales with what the data is worth, not with which chain it happens to sit on. | Security |
Figure 1 — From the problem to the opportunity. → Full detail in Architecture.
The vision. When the marginal cost of recording approaches zero and security spend matches data value, a blockchain can become a verifiable record layer for almost every domain — far beyond finance: identity, supply chain, content, IoT, science.
Stated plainly: XChain is a public good, open source (Apache-2.0), running on a public testnet. Testnet coins have no value — no token sale, no mainnet, not a feature commitment or investment advice. What is live versus in progress is tracked honestly in Status & Live Network.
The argument proceeds the way the design was derived: first the principles that constrain everything (and what the core refuses to do), then the structure those principles force, then the mechanics — cross-chain model, security, primitives — then an honest placement among existing systems, and finally the operational record. Each section closes by naming the question the next one answers. The first question is the deepest: what rules decide where anything belongs?