The research team demonstrates a modular, scalable hardware architecture for a quantum computer

This article has been reviewed according to Science X’s editorial process and policies. The editors have emphasized the following attributes while ensuring credibility of the content:

facts verified

peer reviewed publication

trusted source

to correct


Researchers have developed a modular manufacturing process to produce a quantum system-on-a-chip that integrates a series of qubits of artificial atoms into a semiconductor chip. Credit: Sampson Wilcox and Linsen Li, RLE.

x close


Researchers have developed a modular manufacturing process to produce a quantum system-on-a-chip that integrates a series of qubits of artificial atoms into a semiconductor chip. Credit: Sampson Wilcox and Linsen Li, RLE.

Quantum computers hold the promise of being able to quickly solve extremely complex problems that could take the world’s most powerful supercomputer decades to solve.

However, achieving this performance requires building a system with millions of interconnected building blocks called qubits. Creating and controlling so many qubits in a hardware architecture is a huge challenge that scientists around the world are trying to tackle.

Toward this goal, MIT and MITER scientists have demonstrated a scalable, modular hardware platform that integrates thousands of interconnected qubits into a customized integrated circuit. This “quantum-system-on-chip” (QSoC) architecture allows researchers to precisely tune and control a dense array of qubits. Multiple chips could be connected using an optical network to create a large-scale quantum communication network.

By tuning qubits over 11 frequency channels, this QSoC architecture enables a new proposed “entangled multiplexing” protocol for large-scale quantum computing.

The team spent years perfecting the complex process of making two-dimensional arrays of atom-sized qubit microchips and transferring thousands of them onto a carefully prepared complementary metal-oxide-semiconductor (CMOS) chip. This transfer can be done in a single step.

“We’re going to need a lot of qubits and a lot of control over them to really harness the power of a quantum system and make it useful. We’re designing an entirely new architecture and manufacturing technology that can support the hardware scalability requirements.” system for a quantum computer,” says Linsen Li, a graduate student in electrical engineering and computer science (EECS) and lead author of a paper on the architecture.

Li’s co-authors include Ruonan Han, associate professor at EECS, head of the Terahertz Integrated Electronics Group and member of the Research Laboratory of Electronics (RLE); lead author Dirk Englund, EECS professor, principal investigator of the Quantum Photonics and Artificial Intelligence Group and RLE; as well as others at MIT, Cornell University, Delft Institute of Technology, Army Research Laboratory, and the MITER Corporation. The paper appears in Nature.

Diamond microchips

Although there are many types of qubits, the researchers chose to use diamond color centers because of their scalability advantages. Previously, they used such qubits to make integrated quantum chips with photonic circuits.

Qubits made from diamond colored centers are “artificial atoms” that carry quantum information. Because diamond color centers are semiconductor systems, qubit production is compatible with modern semiconductor manufacturing processes. They are also compact and have relatively long coherence times, which refers to the amount of time the qubit state remains stable, thanks to the clean environment provided by the diamond material.

In addition, diamond color centers have photonic interfaces that allow them to remotely connect or interconnect with other non-adjacent qubits.

“The conventional assumption in the field is that the inhomogeneity of the diamond’s color center is a disadvantage compared to an identical quantum memory such as ions and neutral atoms. However, we turn this challenge into an advantage by embracing the diversity of artificial atoms: Each atom has its own spectral frequency , which allows us to communicate with individual atoms by voltage tuning them into resonance with a laser, similar to tuning the dial on a small radio,” says Englund.

This is particularly difficult because researchers must achieve this on a large scale to compensate for the inhomogeneity of qubits in a large system.

In order to communicate via qubits, they need to have several such “quantum radios” tuned into the same channel. Achieving this condition is almost certain when scaling to thousands of qubits.

To this end, the researchers overcame this challenge by integrating a large array of diamond colored center qubits onto a CMOS chip that provides the control selectors. The chip can be incorporated with built-in digital logic that quickly and automatically reconfigures the voltage, allowing the qubits to achieve full connectivity.

“This compensates for the inhomogeneous nature of the system. With the CMOS platform, we can quickly and dynamically tune all qubit frequencies,” explains Li.

Lock making and release

To build this QSoC, the researchers developed a manufacturing process to transfer “microchips” of diamond color centers onto a large-scale CMOS substrate.

They began by producing a series of diamond colored center microchips from a solid block of diamond. They also designed and fabricated nanoscale optical antennas that enable more efficient free-space collection of photons emitted by these colored center qubits.

They then designed and mapped a chip from a semiconductor foundry. Working in the MIT.nano clean room, they then processed the CMOS chip to add microscale sockets that match the array of diamond microchips.

They created their own transmission setup in the lab and used a lock-and-release process to integrate the two layers by locking diamond microchips into sockets on a CMOS chip. Since the diamond microchips are loosely bonded to the diamond surface, when the diamond volume is released horizontally, the microchips remain in the sockets.

“Because we can control the production of both the diamond and the CMOS chip, we can create a complementary pattern. This way, we can transfer thousands of diamond chips to their respective sockets all at the same time,” says Li.

The researchers demonstrated 500-micron-by-500-micron transmission for an array of 1,024 diamond nanoantennas, but could use larger diamond arrays and a larger CMOS chip to expand the system further. In fact, they found that with more qubits, frequency tuning actually requires less voltage for this architecture.

“In this case, if you have more qubits, our architecture will perform even better,” says Li.

The team tested many nanostructures before determining the ideal microchip array for the lock-and-release process. However, making quantum microchips is not an easy task and the process took years to perfect.

“We iterated and developed the recipe for making these diamond nanostructures in the MIT cleanroom, but it is a very complicated process. Obtaining the diamond quantum microchips required 19 nanofabrication steps, and the steps were not straightforward,” he adds.

In addition to their QSoC, the researchers developed an approach to characterize the system and measure its performance on a large scale. For this purpose, they built their own cryo-optical metrology setup.

Using this technique, they demonstrated an entire chip with more than 4,000 qubits that could be tuned to the same frequency while preserving their spin and optical properties. They also built a digital twin simulation that connects the experiment with digitized modeling, helping them understand the underlying causes of the observed phenomenon and determine how to effectively implement the architecture.

In the future, researchers could increase the performance of their system by improving the materials they used to make the qubits or by developing more precise control processes. They could also apply this architecture to other solid-state quantum systems.

More information:
Dirk Englund, Heterogeneous integration of spin-photon interfaces with a CMOS platform, Nature (2024). DOI: 10.1038/s41586-024-07371-7. www.nature.com/articles/s41586-024-07371-7

Information from the diary:
Nature

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top