Juniper Research projects the PQC market will expand from $1.2 billion in 2026 to $13 billion by 2035 at a 30% CAGR. The catalyst is not a future Q-Day event -- it is the Harvest Now, Decrypt Later tactic, where adversaries are already exfiltrating and stockpiling encrypted data today for future quantum decryption. The window for ignoring this is already closed for any system handling data with a multi-year shelf life.
For hardware engineers, the challenge is concrete. NIST has standardized ML-KEM and CRYSTALS-Dilithium, but deploying lattice-based algorithms on resource-constrained devices -- think microcontrollers, IoT sensors, embedded ASICs -- is not trivial. These algorithms carry heavier compute and memory footprints than RSA or ECC. Hybrid cryptography (layering PQC on top of existing schemes) is the practical migration path, but it roughly doubles key sizes and handshake overhead. That matters when you are targeting a Cortex-M0 or a tiny secure element.
The detail worth noting: the report analyzed adoption across 61 countries, and the clearest gap is at the executive level. Companies are not prioritizing migration because the threat feels distant. The smarter framing -- and the one that maps directly to board-level risk language -- is treating Harvest Now as a current data retention and liability issue, not a future cryptography problem. Hardware teams that want budget for PQC migration projects should be making exactly that argument now.
The caveat: PQC-as-a-service via cloud APIs will work fine for backend infrastructure, but it does nothing for devices that handle cryptography on-chip. That tail of embedded and IoT hardware is where the real engineering work lives, and it will take years to cycle through.