
Sitting in an office at QuEra Computing’s Boston headquarters, Yuval Boger was talking about the recent advancements made in quantum computing that are driving the chorus around an accelerated the timeframe the launch of a usable and reliable system.
“Sometimes it’s hard to see all the amazing progress that’s been happening,” Boger, QuEra’s chief commercial officer, told The Next Platform in a recent interview. “But if you go back a few years – five or ten years ago – the question was, ‘Could people actually build a quantum computer, any quantum computer?’ People understood the science a while ago, or at least the theoretical science, but could you make it? Then IBM and IQM Quantum Computers and Quantinuum and Google and us and many others have said, ‘Yeah.’ By now, it’s a given. People assume that you could build a quantum computer.”
There has been a rush of recent announcements that bolster that contention, with vendors arguing that the question about quantum computers is now “when” rather than “if,” and that “when” could be closer than it appears. Microsoft, Google, and Amazon Web Services all have made high-profile disclosures new or enhanced quantum chips that address the key issue of error correction, with Microsoft declaring that its new Majorana 1 quantum chip means that reliable, fault-tolerant quantum computers are years away, rather than decades.
More recently, IBM in late May announced plans to release a fault-tolerant quantum system – dubbed Quantum Starling – by 2029, that can run quantum circuits with 100 million quantum gates on 200 logical qubits. The next step to Starling will be successive releases of IBM’s upcoming Nighthawk quantum processor starting this year and running through 2028.
The Money Rolls In
There also is more investment money being spend in the market and more partnerships as innovation in the space accelerates. IT giants like Nvidia are looking to muscle their way in, and D-Wave, which has made its annealing Advantage quantum systems available via its Leap cloud platform, sold its first on-premises computer to Germany’s Forschungszentrum Jülich and created a business to sell other systems. That was before it rolled out its Advantage 2 quantum chip and unveiled an aggressive roadmap in March of this year.
QuEra has been in the middle of all this. In February, the company announced a $230 million funding round that included Google, SoftBank Vision Fund 2, and other investors, and followed the $47 million raised in October 2024. The company will use the money to continue building its fault-tolerant technology, hire more scientists and engineers, and grow its partnership lineup of research labs, Fortune 500 companies, and government agencies. In March, QuEra joined Quantinuum and Quantum Machines in announcing their participation in Nvidia’s effort to launch a quantum research center in Boston to develop and provide new technologies and research for quantum computing.
Taking Quantum Systems On The Road
Late last month, QuEra installed its first quantum systems outside of its own labs – where its Aquila system (shown below) is housed – sending the gate-based neutral-atom Gemini quantum computer to Japan’s National Institute of Advanced Industrial Science and Technology. It came a year after the institute contracted with QuEra for $41 million to supply the system to the new G-QuAT quantum-AI research center. The QuEra systems is working next to the Nvidia-powered ABCI-Q supercomputer.
Around the same time, QuEra also delivered a gate-based neutral-atom quantum system to the National Quantum Computing Centre at the Harwell Science and Innovation Campus in Oxfordshire, England.
The pairing of the QuEra Gemini system (below) and Nvidia’s supercomputer in Japan creates a hybrid classical-quantum environment, which QuEra’s Boger said how quantum will likely operate in the coming years.
“A common misconception is that people think that quantum computers will just displace or replace traditional CPUs or GPUs,” he said. “We don’t think that’s the case. It’s just going be one more PU, one more processing unit, in the datacenter. Just like you’re not running Microsoft Word or Zoom on a GPU today, you’re not going to be doing it on a quantum computer. Some things are going to be best for CPUs, some things are going to be best for GPUs, and others for QPUs.”
Taking The Steps
Boger ticked off the points that are leading to quantum computing going from theory and expectation to reality that have been knocked off. In 2023, research by Harvard, QuEra, MIT, NIST, and the University of Maryland showed that qubits could be created to detect and correct errors. Next came the question of whether a quantum system could be built with enough qubits to be useful.
“Based on all this progress, we and others believe that quantum computers can become truly useful in the sense that they solve a business problem that has commercial value in probably two or three years,” he said. “Conventional wisdom is that these first applications will be around chemistry or pharmaceutical or material science and so on.”
The growing investments in recent years indicate that others also believe it. According to The Quantum Insider’s Intelligence Platform, investments in quantum tech in the first quarter surpassed $1.25 billion, more the doubling the amount raised a year ago.
“When you look at valuations of public companies in quantum, you look at the activity from Amazon and IBM and Google and Microsoft, at some point you say it’s unlikely that everyone is wrong,” Boger said. “There’s probably something there. Otherwise, so many serious people, so many serious companies, so much serious money wouldn’t be pouring into this.”
Leveraging Neutral Atom Modality
Part of this is to see in what direction the money goes. There is a range of modalities – different methods for creating and controlling qubits – that companies have latched onto, including superconducting qubits, trapped ions, and photonic qubits. QuEra uses neutral atoms that are held in place by laser beams, which Boger said gives the company advantages in critical areas like cooling and space. QuEra’s systems can run at room temperature – some other modalities need cryogenic cooling or other cooling systems – and it fits in standard 19-inch racks, different from some of the chandelier-style quantum systems. Installing it requires about 20 kilowatts of power.
“We need a little bit of power, stable room temperature just because, if it widely varies, then you get sort of mechanical expansion that could take the system out of alignment,” Boger said. “We want pretty stable temperature. We want air that’s datacenter-clean. We don’t need any condensing humidity or anything like that.”
The atoms themselves are a “beautiful thing because atoms are plentiful and they’re perfect. They’re perfect in the sense that they’re perfectly identical. We can have a million atoms and there are no manufacturing defects. We hold each atom in place with a tiny laser beam. The laser beam acts as what’s called an ‘optical tweezer.’ It just keeps the atom in a place. When you see our system, it has lasers, it has lenses to steer the laser beam in the right direction. It has cameras that photograph the atom. We put the atoms in a vacuum so they don’t interact with just regular air.”
They’re also small, with Boger noting: “Atoms are typically a few microns. A micron is a millionth of a meter. We have atoms that are about four microns apart. If you had 100-by-100 atoms arranged this way – that’s 10,000 atoms on a square – that would be 400 microns. That’s still tiny, tiny, tiny. You need a really high magnification microscope to see it. You need precision lasers to move them around. That’s one of the reasons we’re so optimistic about scaling up, because a million atoms wouldn’t take a lot of space. At some point, it’s moving from a scientific challenge to an engineering challenge.”
From Analog To Digital To Beyond
Since 2022, QuEra’s 256-qubit Aquila quantum system, which is housed in Boston and is based on programmable arrays of neutral Rubidium atoms, has been available via the AWS cloud 130 hours a week. The deployment of the Gemini quantum system in Japan also shows the advancement company is making in its systems.
“The initial Aquila system is an analog system,” Boger said. An analog system is a different way of programming. Japan system is digital system. It’s almost like the difference between a vinyl and a CD for audio. It’s the same use. You can hear music on both, but the way to record, to playback is completely different. It’s a new generation, and as we go to additional generations, we’re going to have more qubits or larger systems. Better qubits in terms of lower error rates or more of these logical qubits, better connectivity with the outside world, and even better interfaces with CPUs and GPUs. The software infrastructure obviously is going to get better and better.”
Great! Sounds like it’s game-on with competitors to be first to demonstrate Quantum Advantage through UFTQC (universal fault-tolerant quantum computing) which requires non-Clifford gate equivalents that successfully circumvent Eastin–Knill’s no-go theorem thus upholding error correction over all computations (eg. {H, S, CNOT, T} or (Clifford + T) gate set). Transversal entangling Cliffords nicely minimize error propagation in lattice surgery but the non-Clifford T’s (and kin) remain a challenge …
I guess that in QuEra’s laser-tweezed neutral Rubidium atoms in vacuum approach (eg. 260-qubit Gemini-Class dynamic array follow-on to Aquila) the plan is for non-Clifford |H⟩ or |T⟩ Wigner-negative entangled contextual magic states to be prepared through n copies of mixed state noisy ancillas (Bravyi & Kitaev, 2004) syndromed over commuting stabilizers, to end up pure and distilled, prior to teleported-injection (in my understanding) as described in their milestone paper from this past December that “opens the door for exploration of hardware efficient generation of quantum magic” ( https://arxiv.org/abs/2412.15165v1 ).
By contrast, in MS’s non-Abelian anyon topological quantum computing Majorana (if it works), the equivalent of “one-bit teleportation” is achieved by an entangling two-body measurement (iiuc) which simplifies matters but requires maintaining a superconducting environment (eg. 30 millikelvin; https://arxiv.org/abs/2502.12252 ), as with Ocelot, Willow, and Starling.
I’m not sure what the “known methods for implementing T gates” are in photonic approaches (Gottesman–Kitaev–Preskill qubits?) but it seems new QLDPC codes at least help with transversal implementation of Clifford gates with 5x to 20x fewer physical qubits than before ( https://photonic.com/news/shyps-codes-announcement/ ).
The competition between research teams and QC modalities is great, and so is the maturing of this tech (eg. QuEra and D-Wave)! It looks to me like we’re on our way to seeing Quantum Advantage demonstrated somewhere relatively soon … (fingers crossed!)
Yes! I too am spellbound by such entangled teleportations, enchanted state superpositions, and charmingly prestidigitated distillations of transversal extra-sensory-perception syndromes … The voodoo is succulent on that one, crunchy on the outside, yet tender and juicy on the inside! ⟨q^8| https://www.science.org/doi/abs/10.1126/science.274.5287.504 |T⟩
It’s by far the best way to celebrate last year’s 100-year anniversary of surrealism, even today, imho ⟨8^p| https://www.artnews.com/list/art-news/artists/surrealism-art-shows-exhibitions-around-the-world-2024-calendar-1234706053/ |H⟩
Especially as that so inspired Heisenberg, Born, Jordan and Dirac to develop their own memorable 1925 non-commutative fish’n’chips Poisson brackets, in preparation for the arrival of Schrödinger’s unique wave of superposed hot-dogs and cats, entangled as they were in savory collapsible balls of quantum wool, relish, and mustard! ⟨q^8| https://www.europhysicsnews.org/articles/epn/pdf/2025/02/epn2025562p15.pdf |8^p⟩
UFTQC Quantum Advantage may arrive by 2030 according to Quantinuum’s trapped-ion roadmap of Helios to Apollo (Sept. 2024 announcement [1]), but two weeks ago IBM stated[2] that its superconducting Starling will get there earlier, in 2029. Great competition, and yes, fingers crossed:
[1] https://www.quantinuum.com/blog/quantinuum-accelerates-the-path-to-universal-fault-tolerant-quantum-computing-supports-microsofts-ai-and-quantum-powered-compute-platform-and-the-path-to-a-quantum-supercomputer
[2] https://www.ibm.com/quantum/blog/large-scale-ftqc