Quantum computing could ease AI's power demands

Artificial intelligence is consuming electricity at a pace that threatens to overwhelm regional power grids, forcing a reevaluation of how the technology is built and powered. As data centers multiply, operators face the prospect of rolling blackouts and strict capacity limits. While experts debate the severity of the crisis, the conversation is shifting from simply generating more power to finding ways to reduce the computational load.

The cost of a growing AI footprint

Grid operators in the United States have already warned they might order rolling blackouts for up to 14 states due to record demand and system shortfalls. The pressure isn’t just theoretical; it is driven by the physical limits of infrastructure and extreme weather events. Meanwhile, the International Energy Agency projects global data-center electricity use could nearly double by 2030, reaching roughly 945 terawatt-hours, a figure comparable to Japan’s total consumption. This surge is largely attributed to AI-optimized facilities, which are expected to quadruple their demand over the same period.

U.S. Department of Energy estimates suggest data centers could consume as much as 12% of the nation’s electricity by 2028. This creates a bottleneck where building new physical infrastructure cannot keep pace with the speed of software advancement. The strain is forcing companies to reconsider where they place these facilities and how they operate them.

Can quantum computing reduce the load?

Some researchers propose a hybrid quantum-classical approach as a potential escape valve. The logic is that quantum processors could handle the most energy-intensive calculations with far less power than traditional silicon chips. Pranav Gokhale, CTO of Infleqtion, argues that these systems manage complex variables efficiently, which could theoretically lower energy consumption. The theory holds that by offloading specific workloads, the overall grid strain would decrease.

This optimism is tempered by practical realities. Quantum computers require extreme cooling to maintain operation near absolute zero, a process that consumes significant energy itself. A modeling study by the National Renewable Energy Lab found that cooling systems often use more power than the computation they support. Marta Estarellas of Qilimanjaro Quantum Tech notes that treating all energy-intensive tasks as a single category is a misconception; quantum systems cannot simply take over every heavy-duty job. The energy balance of these systems remains complex, with the support infrastructure playing a major role in the total equation.