Skip to content

IN DEVELOPMENT · NO CHIP HAS BEEN TAPED OUT

Glossary: root of trust

A root of trust is the part of a system that everything else is checked against. In hardware, it gives a machine a way to prove what it is and what it is running. The first E31 design includes one.

· 2 min read

IN SHORT

A root of trust is the part of a system that everything else is checked against, and that does not itself depend on anything else being trusted first. A hardware root of trust places that part in the chip. In the first BramHA design it is intended to let a machine prove what it is and what it is running. No chip exists.

A root of trust is the part of a system that everything else is checked against. Each check in a system rests on an earlier one, and the root is where that chain begins. Nothing beneath it has to be trusted first.

A hardware root of trust places that part in the chip itself. In general terms it holds a secret that software cannot read out or change, and it acts first when the chip starts.

Why it matters

A machine that joins a network usually identifies itself with something stored in its software: a name, an address, a credential. Software can be copied or replaced, so those claims can be made by something else.

A root of trust in hardware gives a firmer basis. Replacing the machine's software does not replace it. A machine can answer a challenge in a way that only a part holding its secret could. It can also check the software it starts, and report what it started in a form the software cannot forge.

That lets a site ask two questions before it admits a machine: what is this, and what is it running?

How it applies here

Identity is one of the four jobs BramHA is being designed to do. The first design is a real-time controller with a root of trust, so that a machine can prove what it is and what it is running before a site lets it on the network.

The root of trust is in the first design because it is hard to add afterward. It has to be present from the moment a chip starts.

The proof is for the site to use. The silicon is intended to give a site something to check, and the decision to admit a machine stays with the site.

A root of trust concerns identity, not safety. A machine that proves what it is can still be badly programmed, and the safety case remains its vendor's. No security evaluation or safety certification is claimed.

No chip has been taped out, and the specification of the first design is not published. A machine that can prove what it is covers the subject at length, and /chip describes all four jobs.

Questions this raises

Why is it called a root?
Because each check in a system rests on an earlier one, and the chain has to start somewhere. The root is the part that is trusted without a check beneath it.
Does a root of trust decide whether a machine joins a network?
No. It is intended to give the site something to check. The decision to admit a machine stays with the site.
Is the E31 root of trust certified?
No. No chip has been taped out, the specification of the first design is not published, and no security or safety certification is claimed.

KEEP READING