Q Day - the Early Years

The talk argues that the popular picture of a quantum computer suddenly breaking all encryption at midnight is wrong. The speaker expects a slow start with a few machines, most likely run by nation states rather than cloud providers, and attackers who pick targets carefully because each costly run yields only a handful of results. Forgery, not decryption, is presented as the more likely first move.

The hardware estimates are the core evidence. The speaker walks through falling qubit requirements for RSA and ECC, from earlier estimates in the tens of millions down to figures under a million, and newer architectures that trade run time against qubit count. The most probable early targets are signing keys, code-signing pipelines, root certificate authorities, secure boot chains, DNSSEC zones and e-passports, since one forged trust anchor can silently undermine everything beneath it. Bitcoin gets its own section, covering exposed public keys, mempool front-running and the proposed BIP-360 and BIP-361 changes, and the speaker doubts the network will be ready in time.

The closing advice is to expect a few years of forgery and triage before decryption becomes practical, around year three by the speaker's estimate. Organizations should start a cryptographic inventory now, using a CycloneDX bill of materials, and plan for devices that cannot be upgraded to post-quantum algorithms.