The gap just got uncomfortably small

Two research groups have done something that should move quantum computing off the “watch in 2040” list and onto your infrastructure risk register. As Quanta Magazine reported on April 3, 2026, a Caltech team led by Dolev Bluvstein and Madelyn Cain published a design for a quantum computer capable of cracking common encryption with only tens of thousands of qubits — not the million previously considered the minimum bar. Simultaneously, Google announced an implementation of Shor’s algorithm that is ten times more efficient than any prior method. Neither machine exists in deployable form today. But the trajectory has changed.

For context: Shor’s algorithm, devised in the mid-1990s, demonstrated mathematically that quantum hardware could dismantle RSA and elliptic-curve cryptography — the two pillars of almost every TLS handshake, digital signature, and encrypted file transfer in the built-environment supply chain. For three decades, the hardware requirement kept the threat theoretical. That comfort zone is shrinking fast.

←TODAY: Most AEC firms in Switzerland and the DACH region run project data over TLS-encrypted cloud platforms — Autodesk BIM 360, Trimble Connect, BIMcloud — all relying on RSA or ECC key exchange that Shor’s algorithm targets directly.
→3012: In the Zurich-3012 horizon, cryptographic infrastructure is civic infrastructure — buildings negotiate identity, access, and ownership through post-quantum ledgers that today’s architects are not yet specifying.
Fulcrum: The window between “years away” and “decades away” is exactly when procurement decisions get made — and BEPs, data security clauses, and CDE contracts should already be adapting.

Why the hardware suddenly looks plausible

Two converging system trends explain the jump. First: neutral-atom qubits. Unlike superconducting qubits — the locked-in-place transistors that Google and IBM use — neutral atoms can be repositioned in laser arrays. Bluvstein’s prior work at Harvard’s Lukin lab ran sophisticated algorithms on 280 neutral atoms in 2023; a Caltech group led by Manuel Endres separately demonstrated manipulation of 6,100 neutral atoms at once. Mobility matters because the second trend — quantum low-density parity-check (qLDPC) error-correcting codes — requires linking qubits that are physically distant in the array, not just neighbours. Superconducting qubits, fixed in place, cannot do this efficiently. Neutral atoms can.

The combination is the unlock. The surface code, previously the gold standard for error correction, demands thousands of physical qubits to protect a single reliable virtual qubit. qLDPC codes dramatically improve that ratio. Fewer physical qubits per reliable virtual qubit means the total machine size plummets. Bluvstein’s new company, Oratomic, was formed specifically to build toward this target. As Nikolas Breuckmann of the University of Bristol — who was not involved in either paper — told Quanta plainly: “If you care about privacy or you have secrets, then you better start looking for alternatives.”

What this means on a project server this week

AEC firms are not cryptographers, but they are custodians of data that carries long-term sensitivity: land valuations, structural calculations for critical infrastructure, client identity data in BEPs, contractor bids. The relevant question is not “can someone crack our files today” — the answer is still no. The relevant question is “are we storing encrypted data now that will be decryptable in five to ten years?” This is the harvest-now-decrypt-later attack vector that national cybersecurity agencies have been warning about since at least 2022. NIST finalised its first post-quantum cryptographic standards in 2024 (FIPS 203, 204, 205), and the Swiss NCSC has flagged PQC migration as an active priority. Most CDE vendors have not yet published their PQC migration roadmaps.

The system bottleneck is not the quantum computer — it is the update lag in the software stack your projects depend on. Autodesk, Trimble, and Nemetschek group platforms all inherit encryption from underlying cloud infrastructure (AWS, Azure, GCP), and those providers are moving, but slowly and opaquely from a client perspective.

Atelier: If your office is writing a BEP or a CDE protocol for a project with a 10-plus-year data retention requirement — hospitals, infrastructure, public buildings — add a clause requiring the platform vendor to confirm their post-quantum cryptography migration timeline. It is a one-line ask that will separate vendors who have thought about this from those who have not.

The PAZ Five-Beat framing here is not alarmist. It is the same structural habit we apply to materials research or robotic fabrication: when the system curves converge faster than procurement cycles, you want to be one decision ahead, not one report behind. Read the Quanta Magazine piece by Charlie Wood, pull NIST’s PQC standard summary (it is readable in an afternoon), and put the question to whoever manages your office’s CDE contract at the next renewal conversation.

Source: Quanta Magazine