Orchestrating a quantum leap using cold atoms

“Quantum computers will start outperforming supercomputers within the next decade,” predicts Kenji Ohmori, a professor at the Institute for Molecular Science (IMS) in Okazaki, Japan. He sees quantum computers initially coming into their own for quantum chemistry calculations and quantum optimizations, and later solving security-related factoring problems.

The biggest quantum computers constructed to date — such as IBM’s Condor processor that boasts 1,121 qubits — are based on superconductors. But Ohmori and his team are pursuing an alternative approach based on cold atoms, and he sees this method leapfrogging superconductors in the medium term.

“To create a truly useful quantum computer, you need upwards of 10,000 qubits, but a superconducting chip would struggle to hold more than 1,000 qubits,” says Ohmori. “Using cold atoms, it’s going to be easy to scale beyond the 10,000-qubit barrier very soon — we already have an array of 800 optical tweezers to trap up to about 400 cold atoms.”

The cold-atom approach involves using finely focused laser beams (the ‘optical tweezers’) to trap individual atoms in a two-dimensional grid held at temperatures close to absolute zero. Ohmori envisions cold-atom quantum computers making groundbreaking applications in fields ranging from cryptography to modelling complex systems.

Ohmori’s team boasts a core competence that distinguishes them from others in the global cold-atom community. By integrating ultrafast lasers into quantum computing, they have tackled two long-standing challenges that have impeded progress for more than two decades: decoherence and speed.

Dodging decoherence

Decoherence poses a major hurdle to realizing practical quantum computers. It is the process by which quantum states gradually decay back into classical states through interactions with their environment.

One way to sidestep the problem of decoherence is to use timescales that are so short that decoherence has no opportunity to set in. Qubits based on supersized atoms known as Rydberg atoms have this advantage. In these atoms, electrons can be up to a few thousand times farther from the atomic nucleus than in normal atoms. These enlarged atomic states make the atoms more responsive to external fields, making them ideal for precise control in quantum-computing experiments.

In 2016, Ohmori and colleagues demonstrated an intricate method of manipulating and observing quantum states in Rydberg atoms at temperatures near absolute zero¹. They cooled and trapped 87 rubidium atoms using an optical trap and then energized them to high-energy Rydberg states. Using laser pulses that are about 10 picoseconds long (1 picosecond is a trillionth of second, or 10⁻¹² second), the team created an ultracold Rydberg gas and tracked electronic oscillations every femtosecond (10⁻¹⁵ second).

“Picosecond lasers are used for cold-atom computers as they precisely match the movement timescale of Rydberg electrons,” explains Ohmori. “Other quantum computers operating on a slower scale face external-noise issues, but we can forget about decoherence in our ultrafast approach.”

Quantum simulators

Capitalizing on their expertise in ultrafast lasers, Ohmori’s team advanced the field again in 2018 by using picosecond laser pulses with attosecond timing precision to manipulate electronic structure symmetry in atoms². Crucial for realizing quantum control, this research also laid the groundwork for ultrafast quantum simulations, which is promising for studying chemical reactions and charge migration in devices.

“A quantum simulator is an analogue machine. For example, an array of cold atoms could mimic electrons in a solid,” explains Ohmori. “We’ve observed ultrafast electron dynamics in our simulator, such as the rapid development of entanglement within mere hundreds of picoseconds³. Witnessing this kind of evolution was extremely exciting for us.”

In their simulators, Ohmori’s team uses a laser system that projects beams from six angles, forming a cubic lattice of optical traps to hold cold atoms. This intricate setup traps around 10,000 atoms at each lattice point, with distances between atoms being less than 1 micrometre^3,4. This arrangement will facilitate a deeper understanding of collective behaviours in magnetic and superconducting systems.

Racing to the gate

Ohmori’s approach is distinctly digital. Optical tweezers enable each qubit to be precisely addressed and observed, allowing versatile arrays with customizable distances to be created. Such meticulous control is essential for implementing more-advanced quantum algorithms and optimization problems.

A prime illustration of this progress is a 2022 study, in which the team achieved ultrafast energy exchange between two Rydberg atoms, completed within nanoseconds⁵. The experiments successfully observed quantum oscillations and conditional phase shifts — a critical resource for a quantum gate that enables quantum simulation and computation operating at the quantum speed limit set by interactions between dipoles.

“We completed a two-qubit gate in just 6.5 nanoseconds, which is two orders of magnitude faster than any other quantum computer based on cold atoms,” says Ohmori. “This is a disruptive innovation.”

The cold-atom computer’s ability to dynamically move qubits using optical tweezers, allowing researchers to form arbitrary arrays, makes it particularly suitable for optimization problems. This flexibility is key for applications such as designing smart grids and planning wireless networks for agriculture or disaster response. It is a key advantage of cold-atom quantum computers over fixed-qubit systems such as superconducting quantum computers.

The quantum advantage

Ohmori’s advances have ushered in a new era of ultrafast gate operations in quantum computing, but his journey is far from over. A key challenge is overcoming the quantum-scale instability of commercial ultrafast lasers, which are designed for other applications such as mechanical machining and medical surgery.

“To address this challenge, we’ve been developing our own lasers,” says Ohmori. “Our aim is to improve the fluctuation between laser pulses by tens of percent.” Other priorities include enhancing readout speeds and implementing quantum error correction.

With a dearth of talented individuals worldwide, partnering with experienced institutes such as IMS may prove key to advancing in a world seeking to navigate the balance between global collaboration and intellectual-property rights.

“Quantum computing is no longer just lab science; it’s evolved into big science. We’ve formed a substantial, efficient team under our Moonshot Goal 6 programme, collaborating with major partners like Kyoto University, RIKEN and Hitachi, Ltd.,” explains Ohmori. “International collaborations are essential because each country doesn’t necessarily have enough talent. We need to address this issue to develop quantum computers for the benefit of humanity as quickly as we can.”