Artificial intelligence has transformed how companies process data and make decisions—but Silicon Valley’s biggest players are already chasing what could be the next technological breakthrough: quantum computing. Unlike AI, which accelerates existing processes, quantum computing promises to unlock entirely new capabilities, from simulating molecules for drug discovery to solving problems far beyond the reach of today’s fastest supercomputers. The industry is projected to reach $2 trillion by 2035, according to McKinsey.
At Nvidia’s GTC 2025, quantum computing took center stage with a dedicated “Quantum Day,” where experts explored its potential to tackle problems such as weather modeling and drug discovery—challenges that even AI models and hardware-accelerated computers struggle to handle. Major tech players including Microsoft, Amazon, Google, and Nvidia are developing proprietary quantum technology, exploring how it can integrate with AI models to create future-ready infrastructure.
But there’s a key hurdle: scaling qubits.
Qubits—the fundamental units of quantum data—must scale into the thousands for quantum computing to surpass the capabilities of AI. Unlike classical bits, which exist as 0 or 1, qubits can exist in multiple states simultaneously, enabling exponentially faster processing of complex calculations.
To address this, California-based Atom Computing is working alongside Microsoft. In September 2024, Microsoft announced a collaboration with Atom to build the world’s most powerful quantum machine, offering a scalable commercial system available for order. By November 2024, the companies had entangled 24 logical qubits and successfully ran a quantum algorithm using 28 logical qubits. As of 2025, Atom Computing’s neutral atom-based system features 1,180 qubits. But is the technology ready for complex real-world use cases?
“There is no single, concrete number of logical qubits and associated performance metrics that will suddenly unlock every possible application,” Remy Notermans, director of strategic planning at Atom Computing, told Fast Company. “Around 100 logical qubits, certain scientific applications can be explored that will go well beyond classical computing capabilities, and economically valuable applications are expected to become accessible at around 1,000 logical qubits.”
How Atom Computing Stacks Up Against Competitors
Atom Computing’s systems use trapped neutral atoms as qubits—a proprietary approach that allows for precise control. Unlike ionized atoms, neutral atoms retain all their electrons. The company also uses laser cooling and optical tweezers to trap and manipulate individual atoms. Other quantum computing companies—including D-Wave, Phasecraft, Zapata Computing, and Algorithmiq—are also developing infrastructure and algorithms to optimize today’s quantum hardware.
Notermans said Atom’s flagship system currently enables 50 working logical qubits. “With an aggressive roadmap, we anticipate having 100 working logical-qubit and 1,000 logical-qubit systems commercially available in the next few years,” he said. “The high scalability of our neutral atom technology means we have better logical qubits that is significantly faster than other approaches.”
When asked what made Microsoft confident in Atom’s long-term potential over other approaches like superconducting or trapped ion qubits, Atom Computing Chief Product Officer Justin Ging cited scalability and flexibility as key advantages.
“Neutral atom technology enables multiple critical platform capabilities such as high-fidelity gate operations, all-to-all qubit connectivity, long coherence times, and mid-circuit measurement with qubit reset and reuse,” Ging told Fast Company. “A lot of valuable R&D work can be done with Atom’s current systems, which have all the building blocks to construct many logical qubits, allowing researchers to explore error-correction and efficient logical-qubit algorithms for unlocking the first scientifically valuable applications.”
Beyond the challenge of scaling qubits, Ging noted that quantum computing is inherently capital-intensive. “Practical quantum tech should not be held to similar timeline expectations as the software industry,” he said. “Quantum chemistry and materials science applications show a lot of promise for being first to take advantage of quantum computing systems that have around 100 logical qubits.”
Why Investors Are Betting on Quantum-AI Integration
While quantum computing is still evolving toward commercial practicality, big tech is betting on its potential to enhance AI model computation. AI models require vast amounts of energy, but integrating quantum computing could improve efficiency and boost reasoning capabilities.
“Quantum computing will augment AI supercomputers to tackle some of the world’s most important problems, from drug discovery to materials development,” Nvidia CEO Jensen Huang wrote in a recent blog post announcing the company’s Accelerated Quantum Research Center.
One notable breakthrough in the quantum-AI sector is Google’s Willow quantum chip, which solved a random circuit sampling (RCS) benchmark problem in just five minutes—a task that, according to Google, would take the world’s fastest supercomputer 10 septillion years to complete.
“Current AI models are trained on massive datasets that are primarily based on the human experience,” Notermans explained. “If an AI model alone is used to answer a question in problems related to drug discovery, there is going to be a significant uncertainty in its ability to answer the question reliably. Quantum computers can be used alongside to generate quantum-physics-based data that can supplement an AI model’s training dataset, making the overall performance of the AI model more accurate.”
Whether a powerful new paradigm emerges by combining classical and quantum computing to transform AI capabilities remains to be seen.