ESG & Sustainability

ESG & Sustainability#

Measuring and optimizing the environmental impact of quantum computing. Carbon footprint of dilution refrigerators, renewable energy integration, and sustainable quantum infrastructure.

The Energy Footprint of Quantum Computing#

Quantum computers have a unique energy profile compared to classical data centers:

Component

Power Consumption

Notes

Dilution refrigerator

15-25 kW

Cooling to 15 mK; runs 24/7

Control electronics

5-15 kW

Room-temperature microwave generators

Classical co-processing

10-50 kW

Error correction, compilation

Facility overhead

20-40 kW

HVAC, UPS, networking

Total per QPU

50-130 kW

Comparable to a small classical cluster

For comparison, a single NVIDIA H100 GPU consumes ~0.7 kW. A quantum computer consumes 70-180x more power per processing unit, but solves fundamentally different problems.

Sustainability Metrics#

QDiana tracks four ESG metrics for quantum operations:

Energy Efficiency#

\[\text{QE}_{\text{eff}} = \frac{\text{Useful quantum operations (gates)}}{\text{Total energy consumed (kWh)}}\]

Carbon Intensity#

\[\text{CI}_q = \text{Energy (kWh)} \times \text{Grid carbon factor (gCO}_2\text{/kWh)}\]

Norwegian quantum computing benefits from one of the lowest grid carbon factors in Europe (~8 gCO\(_2\)/kWh from hydroelectric power), compared to the EU average (~230 gCO\(_2\)/kWh).

Helium Consumption#

Dilution refrigerators use helium-3 and helium-4. Helium-3 is scarce (produced primarily as a byproduct of nuclear weapons maintenance). Sustainable quantum computing requires closed-cycle helium recovery systems.

Hardware Utilization#

\[\text{Utilization} = \frac{\text{Time executing quantum circuits}}{\text{Total uptime}}\]

Current quantum computers have utilization rates of 5-20%. Improving scheduling and batching can significantly reduce per-computation energy costs.

Cooling Optimization Strategies#

  1. Pulse-level optimization: Shorter gate sequences reduce total refrigeration time per computation

  2. Batch scheduling: Group similar jobs to minimize refrigerator cycling

  3. Hybrid workload splitting: Run only quantum-essential subroutines on the QPU; keep classical work on efficient CPUs/GPUs

  4. Location optimization: Site quantum data centers near renewable energy sources and cold climates

Further Reading#

  • Jaschke, D. and Montangero, S. (2023). “Is quantum computing green? An estimate for an energy-efficiency quantum advantage.” Quantum Science and Technology.

  • Green Quantum Certification – QDiana’s sustainability certification program