Are we ready for quantum computing?

By Jeremy Benjamin | 7th November 2025 | Quantum, Skills, Data Centres

The complexity of quantum isn’t just the qubits — it’s the building

Different quantum modalities impose radically different facility constraints:

  • Superconducting qubits need dilution refrigerators with base temperatures around 10 millikelvin, plus low vibration and tight electromagnetic shielding. That means cryoplants, ultra‑clean signal wiring across multiple temperature stages, and mechanical isolation beyond typical white‑space floors. Commercial cryogenic systems routinely hit ~10 mK at the mixing chamber — orders of magnitude colder than anything in today’s AI halls. cdn.bluefors.com
  • Trapped‑ion systems operate at room temperature but demand ultra‑high vacuum, precision lasers, and exceptional vibrational and thermal stability — again, a very different controls and facilities profile than GPU halls. NQCC
  • Photonic and neutral‑atom approaches change the mix (lasers, optics benches, UHV), but they don’t free you from exacting environmental controls or specialist safety regimes.

If you want a glimpse of “production” quantum facilities, look at IBM Quantum System One installations — purpose‑built enclosures around a cryostat and control stack, installed in enterprise settings like Cleveland Clinic. These systems are not just racks; they are precision instruments that the facility must protect from vibration, EMI, temperature swings, and operator error. Cleveland Clinic+1

Lesson from AI: cooling moved faster than the industry

The AI surge pushed rack power densities from “ambitious” to “non‑negotiable.” Most operators were not ready.

  • In 2024, Uptime Institute found only 22% of organizations had any direct liquid cooling (DLC) deployed, and less than 10% of racks used DLC, even as demand surged — a classic readiness gap. Uptime Institute+1
  • ASHRAE had to push new liquid‑cooling guidance to keep up with device heat flux and reliability risks — the standards were catching up to the market. Data Center Dynamics+1
  • Hyperscalers responded with step‑changes in reference designs: Meta showed a 140 kW liquid‑cooled AI rack at OCP 2024; Google described 1 MW IT racks and a Project Deschutes liquid distribution unit to make facility integration practical. NVIDIA’s Blackwell‑era systems moved to rack‑scale, liquid‑cooled assemblies (NVL72), claiming 25× energy and 300× water efficiency improvements vs. prior air‑cooled setups — again, wholesale architectural shifts, not incremental tweaks. NVIDIA Blog+3Data Center Frontier+3Google Cloud+3

Takeaway: when compute changes fast, facilities, codes, and skills lag. DLC adoption, coolant chemistry, leak detection, CDU integration, water treatment, and materials compatibility (Al/Cu/Glycol/DI) all had to be relearned at speed. Quantum will compress that learning curve further — but with cryogenics, UHV, and precision stability on the critical path.

Why security will accelerate quantum deployment (and budgets)

The “harvest‑now, decrypt‑later” risk is no longer a theoretical talking point. In August 2024, NIST finalized the first PQC standards (FIPS 203 ML‑KEM; FIPS 204 ML‑DSA; FIPS 205 SLH‑DSA), and the NSA’s CNSA 2.0 sets transition expectations for national security systems. The UK NCSC is also telling operators to plan full PQC migration by the mid‑2030s. This is a giant, multi‑year remediation program that primes organizations for serious quantum budgets — for R&D, for pilots, and eventually for on‑prem or near‑prem quantum capacity in regulated sectors. NCSC+4NIST Computer Security Resource Center+4NIST Computer Security Resource Center+4

On the R&D side, the technical signal is improving:

  • Google showed that scaling a surface‑code logical qubit can suppress errors — a pivotal milestone for fault tolerance. Google Research
  • Microsoft + Quantinuum demonstrated highly reliable logical qubits using “qubit virtualization” atop trapped‑ion hardware, reporting 14,000 runs with no observed errors during testing and a dramatic improvement in logical error rates — signs that fault‑tolerance building blocks are maturing. Microsoft Azure+1

Net result: with policy deadlines and technical milestones aligned, the pressure to make quantum operational will rise — and facilities will need to be ready.

The coming skills stack: merge data center, fusion/nuclear, and quantum lab disciplines

Quantum data centers (QDCs) will be hybrid environments: parts of a semiconductor fab, parts of an accelerator hall, and parts of a Tier III/IV data center. Here’s a concrete cross‑disciplinary map.

1) Cryogenics and low‑temperature operations (fusion/nuclear → quantum facilities)

  • Design, operation, and safety of helium and nitrogen systems; recovery loops; purification; oxygen deficiency hazard (ODH) assessment; relief and venting; materials selection at cryogenic temperatures. (Standards like NFPA 55, ASME B31.3, and European cryogenic vessel standards apply.) nfpanorm.com+2allaboutpiping.com+2
  • Practical lessons from ITER: 4 K helium for superconducting magnets, 80 K thermal shields, cryoplant distribution, and commissioning at scale — exactly the kind of plant that informs reliable, maintainable cryo infrastructure. ITER – the way to new energy+1
  • ODH practices from CERN/NIST — ventilation, monitoring, zoning, and procedures — should become table stakes for any site hosting dilution refrigerators. hse.cern+1

2) Precision stability and structural dynamics (fusion/nuclear + DC → quantum)

  • Seismic/structural isolation and micro‑vibration control will matter in ways they never did for air‑cooled racks. ITER’s seismic isolation pit and base‑isolated Tokamak Complex illustrate the kind of modeling and qualification chain (FRS, PSD analysis) we’ll need around quantum rooms and cryostat anchors. ITER – the way to new energy+1
  • Acoustics and EMI: pulse‑tube cryocoolers and control electronics inject mechanical and electromagnetic noise; quantum systems vendors already offer added vibration isolation — facility teams must treat this like a first‑class design variable, not an afterthought. Bluefors.com

3) Advanced cooling integration (DC → quantum)

  • From CDUs and warm‑water loops for GPU racks to helium compressors and cold boxes for cryostats — both require redundancy strategies, corrosion control, bypass modes, and maintainability under Tier objectives. OCP’s Advanced Cooling projects (CDUs, facility integration) are the right launch pad for reference designs that add cryo. Open Compute Project+1

4) Controls, timing, and instrumentation (quantum lab + DC)

  • Integrate quantum control electronics (microwave/laser timing, AWGs, cryo‑CMOS) into BMS/SCADA with deterministic timing, EMI‑aware cable routing, and lifecycle management. Bluefors’ notes on cryogenic measurement infrastructure highlight just how instrument‑heavy a quantum stack is across temperature stages. Bluefors.com

5) Reliability engineering (nuclear/fusion + DC)

  • Apply FMEA/FMECA, configuration control, formal commissioning (FAT/SAT) and permit‑to‑work regimes familiar to nuclear projects — but tuned for quantum rooms co‑located with classical compute. The goal is resilience with serviceability: cryo subsystems must be concurrently maintainable without de‑tuning qubits.

Why bring in fusion/tokamak talent?

Because they’ve already solved adjacent problems at scale:

  • Cryoplant scale‑up and distribution (helium and nitrogen) with complex duty cycles and transients.
  • Base‑isolated, high‑mass structures qualified for seismic events and sensitive equipment.
  • Rigorous safety culture (ODH, pressure systems, vacuum hazards) and systems engineering across thousands of interfaces.

ITER’s public documentation shows exactly the sort of 4 K helium loops, cryoplant capacity and seismic modeling that translate into a robust quantum facility design discipline. Pair that with data center expertise in global rollout, multi‑stakeholder builds, and SLAs, and you have the composite team quantum needs. ITER – the way to new energy+2Fusion for Energy+2

A practical upskilling blueprint (12–24 months)

  1. Standards & safety baseline (all designers/owners).
    • Train teams on NFPA 55, ASME B31.3, and cryogenic vessel standards (EN 13458 family). Add ODH assessment methods and work controls (CERN/NIST/ESS references are excellent primers). Build a cryogenic MOP/SOP/LOTO library. NIST+4nfpanorm.com+4allaboutpiping.com+4
  2. Create a “Quantum Room” reference design.
    • Start with an OCP CDU loop and ASHRAE TC 9.9 liquid‑cooling practices, then extend to a cryogenic annex: compressor room, cold box, vacuum pumps, gas handling and reclaim, ODH detection/venting, EMI shielding, and vibration‑isolated plinths. Make it modular enough to support superconducting or trapped‑ion equipment. Open Compute Project+1
  3. Site selection & civil.
    • Prefer low‑seismic, low‑vibration locations; add base isolation and vibration‑attenuating flooring for rooms housing dilution refrigerators. Use techniques borrowed from tokamak buildings to manage coupled equipment‑structure responses. wcee.nicee.org
  4. Cooling & power integration.
    • For AI and classical HPC nearby, standardize DLC and RDHx playbooks so you’re not chasing bespoke fixes. Treat cryo heat rejection as a first‑class load on your plant, with dedicated redundancy and power quality requirements. (NVIDIA’s NVL72 and hyperscaler projects underscore why this must be “by design,” not “bolt‑on.”) NVIDIA+1
  5. Controls & telemetry.
    • Extend BMS/EPMS to include cryo parameters (pressures, flows, evaporator temps, vacuum levels), EMI/noise sensors, and timing synchronization. Define alarm tiers that reflect qubit health — not just facility state.
  6. Reliability & commissioning.
    • Build nuclear‑style readiness reviews, FMEAs, and SATs into the quantum room hand‑over. Require vendor‑witnessed demos: cool‑down/warm‑up cycles, ODH drills, loss‑of‑utility scenarios, and cryostat service windows that avoid qubit re‑tuning.
  7. Security alignment.

What a “quantum‑ready” skills profile looks like

For data center designers/owners:

  • DLC/RDHx/CDU expertise; OCP Advanced Cooling patterns. Open Compute Project
  • Cryogenic fundamentals: helium/nitrogen properties, ODH mitigation, materials at cryo temperatures, pressure relief and venting. NIST
  • Structural dynamics: floor response spectra, base isolation options for sensitive rooms. wcee.nicee.org
  • EMI/acoustics hygiene: cable routing, shielding, compressor isolation, pulse‑tube mitigation. Bluefors.com

For fusion/nuclear engineers moving in:

  • Tiered uptime concepts, concurrent maintainability in occupied facilities, global supply chains, and multi‑tenant constraints.
  • Data‑center commissioning culture (ITIL/DevOps meets MOP/SOP).
  • Cyber and compliance (NERC‑like thinking applied to PQC and control networks).

For quantum hardware teams:

  • “Design‑for‑facility” documentation: allowable vibration spectra, EMI budgets, exhaust/venting, helium reclaim interfaces, and planned maintenance rhythms.
  • Control stack integration requirements for BMS/EPMS visibility and alarms.

What about timelines and “when”?

Governments are setting the pace on security, not compute. With PQC standards finalized in 2024 and national guidance converging on 2030–2035 migration horizons, boards now have dates on the wall — and the budget lines follow. That gives data center owners a clear mandate: prepare facilities and skills before quantum hardware maturity forces the same scramble AI did for cooling. NIST Computer Security Resource Center+2NSA+2

The bottom line

Quantum will not fit neatly into today’s AI data halls. It will merge fusion‑grade cryogenics and seismic discipline with cloud‑grade resiliency and rollout. The AI era proved that when compute leaps, facilities and people must leap with it — liquid cooling’s slow readiness in 2023–2024 is Exhibit A. If we upskill data center designers now, recruit from the fusion/nuclear skill market, and codify a shared standards playbook, we can make quantum deployments reliable, maintainable, and globally repeatable.

Design leaders who start this journey today won’t just be “quantum‑ready.” They’ll be the ones writing the template everyone else adopts.

Selected sources

Millikelvin operations: Bluefors dilution refrigerator application notes. cdn.bluefors.com

Liquid cooling & readiness: Uptime Institute Cooling Survey 2024; ASHRAE TC 9.9 guidance; OCP cooling/CDU materials; Meta/Google/NVIDIA OCP updates. NVIDIA+6Uptime Institute+6Uptime Intelligence+6

Quantum milestones: Google error‑suppression milestone; Microsoft+Quantinuum reliable logical qubits. Google Research+2Microsoft Azure+2

Security policy: NIST FIPS 203/204/205; NSA CNSA 2.0; UK NCSC PQC timelines. NCSC+4NIST Computer Security Resource Center+4NIST Computer Security Resource Center+4

Cryo & seismic (fusion): ITER cryogenics and seismic isolation references. ITER – the way to new energy+2ITER – the way to new energy+2

Cryogen safety/ODH: CERN/NIST guidelines. hse.cern+1

Scroll to Top