A groundbreaking advancement in quantum computing has emerged that could dramatically accelerate artificial intelligence capabilities and applications.
Researchers at Chalmers University of Technology in Sweden have developed a highly efficient amplifier that represents "the most sensitive amplifier that can be built today using transistors." The team has managed to reduce its power consumption to just one-tenth of that required by today's best amplifiers without compromising performance.
The innovation comes from a smart design that only switches on when reading data from qubits. This reduced power consumption helps minimize interference with the qubits and could enable the construction of larger, more powerful quantum computers. Reading quantum information is extremely delicate—even slight temperature fluctuations, noise, or electromagnetic interference can cause qubits to lose their quantum state. Since amplifiers generate heat that causes decoherence, researchers have been pursuing more efficient qubit amplifiers.
Unlike other low-noise amplifiers, the new device is pulse-operated, activating only when needed for qubit amplification rather than remaining continuously on. Since quantum information is transmitted in pulses, a key challenge was ensuring the amplifier activated rapidly enough to keep pace with qubit readout. The Chalmers team addressed this by using genetic programming to enable smart control of the amplifier, allowing it to respond to incoming qubit pulses in just 35 nanoseconds.
This advancement is essential for scaling up quantum computers to accommodate significantly more qubits. As the number of qubits increases, so does the computer's computational power and capacity to handle highly complex calculations. However, larger quantum systems require more amplifiers, leading to greater power consumption that can cause qubit decoherence. "This study offers a solution in future upscaling of quantum computers where the heat generated by these qubit amplifiers poses a major limiting factor," says Jan Grahn, professor of microwave electronics at Chalmers.
The breakthrough coincides with recent research demonstrating that even small-scale quantum computers can enhance machine learning performance using novel photonic quantum circuits. These findings suggest that today's quantum technology isn't just experimental—it can already outperform classical systems in specific tasks.
Quantum computers have the potential to tackle problems far beyond the reach of today's most powerful machines, opening doors in drug discovery, cybersecurity, artificial intelligence, and logistics. The ultra-efficient amplifier developed at Chalmers only switches on when it's time to read data from qubits. Thanks to its smart, pulse-based design, it uses just a tenth of the power required by current top-tier models.
Many current large language models require over 1 million GPU hours to train, while quantum neural networks promise more efficient processing of complex, high-dimensional datasets compared to classical neural networks. Beyond speed improvements, quantum computing could revolutionize AI through enhanced optimization algorithms, more sophisticated model simulations, and significantly reduced energy consumption for training AI models.
"We expect the first significant breakthroughs in Quantum AI to emerge by the end of this decade and the beginning of the next, as we transition from today's noisy quantum devices to error-corrected quantum computers with tens to hundreds of logical qubits," explains Dr. Ines de Vega, Head of Quantum Innovation at IQM. "These machines will allow us to move beyond purely experimental NISQ quantum algorithms, unlocking practical and potentially unexpected advantages for AI applications. The fusion of Quantum Computing and AI has the potential for a massive impact on the world. Quantum and AI together could solve problems that classical computers cannot, making AI more efficient, faster, and more powerful."