Anatomy of a Quantum Block

Every blockchain is, at heart, a chain of blocks. What makes a QBLEX block different is what lives inside its header: not proof of wasted electricity or staked capital, but a proof of coherence — a measurement record from real quantum hardware that no classical system could have produced. Let's open one up.

A signature made of physics

A conventional block header commits to its contents with a cryptographic hash and, in Proof-of-Work, a nonce that took enormous computation to find. A QBLEX header commits to something a computer cannot fake: a transcript of quantum measurements whose correlations violate a Bell inequality.

Concretely, each block carries a Mermin (GHZ) measurement record — a batch of shots from a three-qubit entangled state, measured on gate-based quantum hardware (our runs come from Rigetti's Cepheus-1-108Q). The block's "signature" isn't a number someone chose; it's what the world returned when we measured entangled particles.

The fence: beating the classical bound

From those shots we compute a single statistic, M. For a Mermin test, any classical system — anything without genuine entanglement — is mathematically confined to |M| ≤ 2. Quantum mechanics can push it higher.

A real block clears that fence. Our published runs land at |M| = 2.685, 2.377, and 2.724 — comfortably past 2, where only quantum correlations can reach, and never at the idealized maximum of 4. That last point matters: real hardware has noise, so a perfect 4 would be the tell of fabricated data. An imperfect number that still clears the bound is the honest signature of physics happening on a real device.

The measurement record, committed

Beating the bound isn't enough on its own; you also have to prove the shots shown are the shots that were committed. So every individual ±1 outcome is folded into a Merkle tree, and the header stores its root, the runs root. Flip a single measurement in any of the 256 shots and that root changes. The record is tamper-evident down to one bit.

Settings you can't cherry-pick

A skeptic's next move is "you chose favorable measurement settings." A QBLEX block closes that door. The measurement bases are fixed by a challenge derived from the block's own position in the chain — its previous head plus a per-block nonce — not chosen freely. Those settings get their own Merkle root, and the verifier recomputes the whole derivation. You can't pick the settings after seeing the data.

The anchor: one hash to bind them

Finally, all of it — device identifier, program hash, both Merkle roots, the challenge, timestamps, and the hardware task records — is folded into a single entry hash. That's the block's anchor: a compact commitment that pins the quantum measurement to specific, third-party-checkable quantum jobs (in our case, AWS Braket tasks on named Rigetti hardware). The provenance doesn't rest on trusting us; it rests on records in Amazon's systems.

Validation: don't trust us, run the math

Here's the part we're proudest of. Every check above — recompute the statistic, rebuild both Merkle roots, re-derive the settings, match the anchor — runs in exact arithmetic, in your own browser, with no call to our servers. A consensus mechanism you can't independently verify isn't consensus at all.

You don't have to take our word for any of this. Verify a real block yourself →