Static vs. Dynamic Root of Trust

A Static Root of Trust begins at boot and builds a measurement chain from the first trusted code through each later firmware component. In this model, every stage measures the next stage before handing over control, so the final integrity report reflects the whole path that executed since reset.

The strength of this approach is that it is straightforward: trust starts early and follows the boot sequence. Its weakness is that the trusted computing base can become very large, because bootloaders, platform initialization code, device firmware, and run-time firmware may all sit inside the chain of trust.

A Dynamic Root of Trust starts later, after boot-time software has already run, and uses a protected launch point to measure only the software that should be trusted from that moment forward. This makes it possible to exclude earlier boot firmware from the trusted computing base, but only if the dynamic launch mechanism is isolated from the untrusted code that invokes it.