TECH

Fujitsu introduced microscopic defects into diamonds to create a new type of quantum computer

Forget traditional quantum chips for a moment. Fujitsu has just unveiled a prototype that uses diamonds, tin atoms, and light particles to tackle one of quantum computing's biggest hurdles: scaling up machine size without losing control of the qubits. While it still requires operating temperatures of an impressive -271.6°C, it features a characteristic that could completely transform how quantum computers are built.

The new computer developed by Fujitsu employs an approach known as "diamond spin."

This doesn't simply mean swapping silicon for a gemstone.

Researchers work with synthetic diamonds containing microscopic defects carefully engineered into their crystal structure. The prototype utilizes tin-vacancy centers, known by the acronym SnV.

In these centers, a tin atom occupies a position between two vacancies within the diamond's structure. These tiny defects allow for the creation and control of quantum states that function as qubits.

Similar systems traditionally use nitrogen-based NV centers. Fujitsu opted for tin because its structure offers greater symmetry and is less susceptible to certain types of external noise.

SnV centers can also emit light with an intensity roughly ten times greater than the conventional approach—an advantage that is particularly appealing given the company's future plans.

However, there is a detail that prevents any conclusion that Fujitsu has solved the issue of extreme temperatures in quantum computing.

The prototype operates at approximately -271.6°C, or about 1.55 Kelvin.

That is just a few degrees above absolute zero.

Even so, it is a higher temperature than the benchmark of approximately -273.13°C cited by Fujitsu for superconducting quantum computers.

Therefore, this new development is not a room-temperature quantum computer.

Far from it. The great promise lies in another feature: the modules can be connected using light.

And this aims to solve a fundamental problem.

Building a few qubits is one thing. Creating thousands or millions of them that work reliably within the same system is far more complex.

Instead of trying to fit everything into a single, gigantic structure, Fujitsu is betting on smaller modules that can be interconnected.

Light can transform small modules into a massive machine...The architecture uses photons to establish quantum connections between modules.

These links can occur between qubits located on different chips and even between systems housed in separate cryogenic units.

To make this possible, Fujitsu developed photonic circuits that combine nanometric diamond structures with alumina optical waveguides capable of carrying the light emitted by SnV centers.

It is like building a giant computer by assembling smaller parts, but using the quantum properties of light to enable them to work together.

This modular strategy could greatly facilitate future expansion.

Fujitsu and the QuTech institute at Delft University of Technology have previously demonstrated operations involving entanglement and quantum gates between NV centers housed in separate cryogenic units.

Now, the company aims to scale this concept up to a much larger architecture.

The 1,000-qubit milestone is still a few years away...This is where an important detail comes in. The computer unveiled now does not feature 1,000 logical qubits.

That figure is part of Fujitsu's technology roadmap.

The company plans to unveil a multi-module diamond-spin computer prototype in 2027. Subsequently, its goal is to achieve a system with 250 logical qubits in the 2030 fiscal year and reach 1,000 logical qubits in the 2035 fiscal year.

The word "logical" is also significant.

Physical qubits are extremely sensitive to noise and errors. Therefore, fault-tolerant quantum computers must combine multiple physical qubits to produce more reliable logical qubits. The hope is that the characteristics of diamond spins will allow for the formation of these logical qubits using fewer physical qubits than some competing approaches.

Fujitsu does not intend to abandon superconducting quantum computers.

Quite the opposite.

The company plans to develop technologies capable of integrating diamond-based systems with superconducting machines, leveraging the advantages of different architectures.

The current prototype has also already been operated in a test environment using the company's hybrid quantum computing platform.

There is still a vast gap between demonstrating a prototype and building a fault-tolerant quantum machine capable of solving commercially relevant problems.

But the experiment points to a possible way to overcome a fundamental obstacle: how to keep adding qubits without turning the computer into an uncontrollable structure.

Fujitsu's answer may lie in a rather unlikely combination.

Tiny defects within diamonds create the qubits. And particles of light may be responsible for linking them all together.

Comment from Vivek Mahajan, Corporate Executive Officer, Corporate Vice President, CTO, in charge of System Platform, Fujitsu Limited

"The diamond-spin approach we have applied in this prototype not only offers exceptional scalability in its own right, but also has the potential to be integrated with superconducting quantum computers to further extend their capabilities, enabling more complex and large-scale computations.

Under our roadmap to achieve a 250 logical qubit system by fiscal 2030 and a 1,000 logical qubit system by fiscal 2035, Fujitsu will continue advancing practical quantum computing across a broad range of areas, from software to hardware, while leveraging the key advantages of the diamond-spin approach, including high fidelity and optical connectivity."

Comment from Dr. Kees Eijkel, General Director, QuTech, Delft University of Technology...“We are delighted to announce this prototype diamond spin quantum computer as a result of the collaborative research conducted since 2020 between Fujitsu, Delft University of Technology, and QuTech. It is a major milestone in our strong collaboration. Demonstrating the scalability expected of diamond spin quantum computing remains a long and challenging journey. However, by further strengthening our collaboration with Fujitsu, we are committed to tackling this ambitious and meaningful challenge and leading the development of next-generation quantum technologies.”

Overview of developed technologies...The prototype features these three technologies developed by Fujitsu:

Heterogeneous material bonding and thinning technology for scalable quantum computing chips...To create quantum computing chips using SnV centers, Fujitsu developed heterogeneous material bonding technology to bond high-quality diamond substrates ion-implanted with tin to alumina/silicon dioxide substrates. Fujitsu also developed thinning technology to reduce the thickness of diamond substrates from several hundred micrometers to several hundred nanometers, making them suitable for use in quantum computing chips.

Photonics-integrated circuit fabrication technology for SnV centers...Fujitsu developed technology to fabricate photonics integrated circuits [3] that integrate nanometer-sized diamond crystals containing SnV centers with alumina optical waveguides, which are transparent in the visible light region, to extract single photons emitted from SnV centers during qubit readout. For diamond processing, Fujitsu utilized the results of joint research with The University of Tokyo.

Quantum circuit conversion technology for diamond spin approach...The diamond-spin approach requires qubit control by combining light, microwaves, and radio frequency waves. Fujitsu developed a mechanism to convert quantum circuits described by quantum gates into control sequence for these physical operations for the diamond spin approach, enabling control from the Fujitsu’s hybrid quantum computing platform.

mundophone