Quantum computing: This new 100-qubit processor is built with atoms cooled down near to absolute zero | ZDNet

By cooling atoms down to near absolute zero and then controlling them with lasers, a company has successfully created a 100-qubit quantum processor that compares to the systems developed by leading quantum players to date. 

ColdQuanta, a US-based company that specializes in the manipulation of cold atoms, unveiled the new quantum processor unit, which will form the basis of the company’s 100-qubit gate-based quantum computer, code-named Hilbert, launching later this year after final tuning and optimization work. 

There are various different approaches to quantum computing, and among those that have risen to prominence in the last few years feature superconducting systems, trapped ions, photonic quantum computers and even silicon spin qubits. 

SEE: Building the bionic brain (free PDF) (TechRepublic)

Cold atoms, on the other hand, haven’t made waves in the quantum ecosystem so far. ColdQuanta’s 100-qubit quantum processor, however, could seemingly compete against the industry’s highest standards: for example, IBM’s current quantum system, Hummingbird, supports 65 qubits.  

And in the next three years, ColdQuanta is hoping to create a system surpassing 1,000 qubits. This again aligns with IBM’s roadmap for quantum hardware, which should see the company releasing a 1,121-qubit quantum computer in 2023. 

“We hear a lot about superconducting and trapped ions and in some respects cold atom is the new kid on the block, but we believe it has great promise in terms of scalability,” Paul Lipman, president of quantum computing at ColdQuanta, tells ZDNet. 

ColdQuanta’s approach consists of treating atoms like qubits, and bringing them down to extremely cold temperatures, where their quantum properties can be manipulated with great precision. This is because, in such an isolated environment, atoms are protected from environmental noise and can retain their quantum properties for much longer.  

Cooling down particles to exert better control over them is not new to the quantum world: Google and IBM’s superconducting processors also require placing qubits in huge dilution refrigerators, where temperatures are brought down to zero kelvin (-273.15C).  

But ColdQuanta’s cold atoms approach goes one step further. Atoms are cooled down to the microkelvin level – that is, a thousand times colder than in the superconducting method. 

Rather than using large refrigerators, however, ColdQuanta traps the atoms with lasers to cool them down, before using a combination of lasers and microwave pulses to arrange them into the gates that make up a quantum circuit.  

“Because we cool them down with lasers rather than dilution refrigerators, we don’t have the same scaling challenges in terms of building enormous fridges that can hold large numbers of qubits,” says Lipman. “We cool them down to microkelvin, but we do that in a device that can fit in your hand at room temperature.” 

What’s more: atoms are ten-thousand times smaller than superconducting qubits, according to Lipman, meaning that many cold atom qubits can be packed closely together on a much smaller space. What would require square-meters worth of space for a superconducting quantum processor can sit on a cold atom system the size of a nail, according to the company. 

“Cold atoms have this intrinsic scalability that is very attractive,” argues Lipman. 

Cold atoms’ ability to scale rapidly is one of ColdQuanta’s key selling points, but there remain some engineering challenges that, for now, still limit Hilbert’s size. The company’s scientists are looking at how the use of lasers changes when the qubit count increases by orders of magnitude, for instance, and testbeds are already underway in the lab to determine the best path forward. 

The fundamental principles of the approach, however, are tested and proven, says Lipman, and cold atoms already perform similarly to leading-edge quantum processors. Not only on qubit count: the company’s data also shows that the system is comparable to IBM and Google’s quantum computers when it comes to connectivity, which refers to the number of qubits that can interact with one another, and coherence, which is the duration of time that quantum properties can be maintained. 

On fidelity, however, the processor lags slightly behind the devices developed by competitors, meaning that the accuracy of ColdQuanta’s system isn’t as high. But part of the optimization work going on now, says Lipman, is dedicated to boosting Hilbert’s performance on fidelity. 

Lipman is confident that these promising results will set ColdQuanta apart in an ecosystem that is growing at pace. New milestones are announced by quantum companies large and small at a rapid pace, and the number of approaches to quantum computing is multiplying fast, each with their own benefits and challenges – making it increasingly difficult to distinguish hype from reality. 

“It’s too early to tell which modality will win the race,” admits Lipman. “If you roll the clock forward two or three years, there might even be modalities that we don’t even have publicly available information on today, but may come to the forefront.” 

“We’ll learn more once the computer is released, but our focus now is to work with potential customers to deliver tangible near-term value.” 

ColdQuanta has not publicly announced any customers yet, but the company is working particularly on optimization problems, which could find applications in logistics, material science and telecommunications. 

The firm also has a long-standing partnership with the Defense Advanced Research Projects Agency (DARPA), which awarded ColdQuanta a total $7.4 million to develop a scalable cold-atom-based quantum computer for defense applications such as resource allocation, logistics, and image recognition.  

Hilbert is expected to launch later this year and will be available over ColdQuanta’s private cloud. The company is also in talks with Amazon, Microsoft and Google to eventually make the quantum computer accessible over AWS, Azure and Google Cloud.