Skip to content

What is Locally Verified Entropy?

Locally Verified Entropy means the randomness is generated inside the trust boundary that consumes it, and the source reports evidence about its own health while it runs. Quantum physics is how the entropy is made; local and verified are the two properties a customer actually buys.

Lead with the second half

Most quantum randomness messaging leads with the quantum. That is the least useful part for a buyer, because "quantum" describes the mechanism rather than the benefit, and every competitor says it too.

What changes an architecture is that entropy is generated where it is consumed, and that the source can say something about its own condition. Those are the two words that survive translation into a requirement document.

Local

Entropy generated inside the boundary that uses it does not have to be trusted across a boundary, transported, or authenticated in transit. It has no dependency on a network, and it does not become unavailable when something else does.

This is possible now because the source became a semiconductor part rather than an instrument — integration is what made locality affordable. The alternative is worth arguing on the merits rather than dismissing.

Verified

The source measures the physical quantities its entropy model depends on and reports them. Not a certificate from evaluation time — a live statement that the assumptions still hold.

That is what makes the entropy claim current rather than historical, and it is the property that lets a system answer questions about a specific key rather than about a product line. Assurance is the general form of this.

What it lets you say

A system with locally verified entropy can state, for a given key, that it was generated from a source inside the boundary, and that the source's health measurements were within their modelled range at the time. Almost no deployed system can say either thing today.

The second sentence is the interesting one commercially, because it is the beginning of key provenance — a question the market does not yet know to ask.