A Canadian startup called Xanadu has developed a new quantum computer named Aurora, which is designed to be easily scalable to meet the computational demands of various scientific challenges, including drug discovery and energy-efficient machine learning. Aurora operates as a “photonic” quantum computer, utilizing photonic qubits—information encoded in light. This process involves the combination and recombination of laser beams across multiple chips using lenses, fibers, and other optical components, depending on the algorithm being executed. The output of an algorithm corresponds to the final number of photons in each laser beam, distinguishing it from the approach used by companies like Google and IBM that encode information in superconducting circuits.

The design of Aurora is modular, comprising four similar units housed in a standard server rack that is slightly taller and wider than an average human. According to Christian Weedbrook, the CEO and founder of Xanadu, to create a functional quantum computer, one would need to replicate and interconnect thousands of these units. This vision contrasts with the traditional view of quantum computing as a specialized chip embedded within a supercomputer, akin to a GPU.

While Xanadu’s current setup includes 12 qubits, which may appear minimal compared to IBM’s 1,121 qubits, Tiwari argues that the quantity of qubits is more indicative of investment levels than a reflection of the technology’s potential. The next objective for Xanadu is to enhance the quality of the photons used in their system, which would simplify error correction procedures. Tiwari notes that during the transmission of lasers through various media—whether free space, chips, or fiber optics—some information is inevitably lost. By reducing this loss, Xanadu aims to minimize errors from the outset. The company is targeting the development of a quantum data center by 2029, with plans for thousands of servers containing a million qubits.