QuEra Computing: The Neutral-Atom Bet to Build a Fault-Tolerant Quantum Computer

QuEra is a Boston quantum computing company that builds machines out of individual atoms held in place by lasers.

It was spun out of Harvard and MIT in 2018, and its scientific pedigree is arguably the strongest in its field.

Its bet is that "neutral atoms" are the most scalable path to a fault-tolerant quantum computer, the kind that could one day solve problems no classical machine can touch.

In February 2025 that bet drew a striking vote of confidence: a $230 million round backed by Google and SoftBank, with NVIDIA joining later.

Key takeaways

• The science is real and peer-reviewed. QuEra's Harvard partners ran algorithms on 48 logical qubits, a landmark in quantum error correction.
• One real product ships today. Its 256-qubit Aquila machine is available on Amazon Braket. Almost everything beyond that is roadmap.
• The backers are blue-chip. Google, SoftBank, and NVIDIA all invested, which is notable because Google is also a direct competitor.
• It is under-capitalized for the endgame. QuEra has raised roughly $250 million, a fraction of what rivals like Quantinuum and PsiQuantum command.
• The timeline is long and uncertain. Its own fault-tolerant milestones sit in 2028 and beyond, and the whole industry has a history of slipping such dates.

Background and origins

QuEra was founded in 2018 to commercialize decades of atomic-physics research from two neighboring labs in Cambridge.

Its scientific founders are a who's who of the field, including Harvard's Mikhail Lukin and Markus Greiner and MIT's Vladan Vuletic and Dirk Englund.

The company's foundational result was a 2021 Harvard demonstration of a 256-atom programmable quantum simulator, the machine that became Aquila.

Alex Keesling, a co-inventor of the core technology, served as founding CEO and now focuses on the technology and production side.

Leadership

Since 2024 the CEO has been Andy Ory, a repeat entrepreneur rather than a physicist.

He previously founded Acme Packet, which he sold to Oracle for about $2 billion, and 128 Technology, acquired by Juniper Networks.

The signal is deliberate: QuEra brought in an operator to turn Nobel-adjacent science into a shipping business, with the labs still driving the physics.

QuEra has raised roughly $250 million since inception, modest by the standards of this capital-hungry field.

It emerged from stealth in 2021 with a $17 million round led by Rakuten, then went quiet on fundraising for several years.

The breakthrough came in February 2025 with a $230 million convertible note led by Google, alongside SoftBank Vision Fund 2 and Valor Equity Partners.

In September 2025, NVIDIA's venture arm joined an expansion of that same round.

Note the structure. The 2025 deal was a convertible note, not a priced equity round, and about $60 million of the $230 million was contingent on conditions.

QuEra has never disclosed a valuation. Secondary-market trackers estimate it in the mid-hundreds of millions, but treat that as an unofficial guess, not a fact.

The technology: computing with single atoms

QuEra's machines store each qubit in a single neutral atom of rubidium, held in a vacuum and pinned in place by a focused laser beam called an optical tweezer.

To make two qubits interact, the atoms are briefly excited into "Rydberg" states, where they balloon in size and push on their neighbors.

This effect, the Rydberg blockade, is the native source of entanglement, and it is what lets QuEra build quantum logic out of atoms.

The approach has three genuine advantages. Every atom is perfectly identical, so there is no manufacturing variation. Coherence times are long, over a second. And because the tweezers can move, the atoms can be rearranged mid-computation, giving flexible connectivity that is a major asset for error correction.

Two machines, two modes

Today's product, Aquila, runs in analog mode, acting as a programmable simulator of physical systems rather than a general-purpose computer.

That makes it well suited to optimization problems and quantum-physics simulation, but it is not a universal gate-based machine.

QuEra's second-generation Gemini class is gate-based, the mode needed for arbitrary algorithms and full error correction, and it is the foundation for everything on the roadmap.

The catch

Neutral-atom gates are slow, on the order of a hundred to a thousand times slower than superconducting gates, a point critics inside IBM have pressed.

Atoms also fall out of their traps and have to be reloaded, and measuring one qubit without disturbing its neighbors is genuinely hard.

QuEra argues that massive parallelism makes up for the slow clock, but that trade-off has not yet been independently benchmarked on commercially useful problems.

The one machine anyone can use today is Aquila, reachable through Amazon Braket since 2022.

Pricing follows Braket's standard model, so it is cheap to experiment with and it is the same whether you are a student or a bank.

In practice a typical 1,000-shot job costs about $10, per the Amazon Braket pricing page, though rates can change so verify before quoting.

On the software side, QuEra maintains Bloqade, an open-source toolkit for programming and simulating neutral-atom machines, plus integration with NVIDIA's CUDA-Q.

It has also started selling on-premises machines, most notably a gate-based system installed in Japan, the company's first deployment outside its own lab.

QuEra's traction is a mix of government programs, HPC deployments, and enterprise pilots rather than large commercial revenue.

It was selected for Stage B of DARPA's Quantum Benchmarking Initiative, worth up to $15 million, a credibility marker in the field.

Its flagship deployment is a roughly $40 million on-premises system for Japan's AIST, installed next to an NVIDIA-powered supercomputer, and it is building a testbed for the UK's National Quantum Computing Centre.

On the enterprise side, it ran a clinical-trial-prediction study with Merck, Amgen, and Deloitte, framed as co-design research rather than proven quantum advantage.

QuEra competes on two levels at once: against fellow neutral-atom startups, and against entirely different quantum technologies.

The market context is enormous but distant. BCG projects quantum could create $450 billion to $850 billion in value by 2040, while the near-term hardware market is only a couple of billion dollars.

Within neutral atoms, Atom Computing is the sharpest threat, since it has Microsoft's distribution and recently out-raised QuEra.

Against other modalities, trapped-ion players lead on qubit quality and revenue, while Google's Willow chip produced the cleanest proof yet that error correction improves as you add qubits.

QuEra's edge is its scaling story and its error-correction pedigree, not its bank balance.

QuEra called 2025 its "year of fault tolerance," and the label was earned in the lab.

Working with Harvard and MIT, its ecosystem published a run of results: logical magic-state distillation, a 3,000-atom array run continuously for over two hours, and a 448-atom machine operating below the crucial error-correction threshold.

These are genuine, peer-reviewed advances, and together they tick off several ingredients needed for a fault-tolerant computer.

There is an honest caveat, though. Most of these headline results are led by the Harvard and MIT academic labs on research apparatus, not features of a product you can rent today.

The most intriguing twist is strategic. Google is both an investor in QuEra and, through its superconducting program, one of its competitors.

That a rival would hedge by backing QuEra's atom-based approach is one of the strongest outside signals that the neutral-atom bet is credible.

What comes next

QuEra's roadmap points to a machine called Libra, a fault-tolerant system it aims to launch on Amazon Braket by 2028.

Beyond that, a 2026 roadmap promises a "gigaquop" system with over a thousand logical qubits toward the end of the decade.

Both are projections, and the entire industry has a track record of missing such dates.

Strengths

• Arguably the strongest scientific pedigree in neutral atoms, via its Harvard and MIT partnership.
• A scaling story built on identical qubits, long coherence, and reconfigurable connectivity.
• Blue-chip validation from Google, SoftBank, and NVIDIA, plus a DARPA selection.
• A live, low-cost product on Amazon Braket that has underpinned dozens of research papers.

Weaknesses and risks

• Under-capitalized. About $250 million is small next to Quantinuum, PsiQuantum, and public rivals, and fault tolerance is a capital war.
• Slow gates and atom loss. These are real physics limits, not marketing quibbles, and the workarounds are unproven at scale.
• Thin commercial revenue. QuEra claims record revenue but discloses no figures, and IonQ dwarfs it on that measure.
• Product versus lab. Its most impressive milestones are academic demonstrations, while its shipping product is still pre-fault-tolerant.
• Long, uncertain timelines. Useful fault tolerance is a late-2020s or 2030s prospect, and hype cycles can dry up funding before then.

QuEra is one of the most scientifically credible companies in quantum computing, and its neutral-atom approach has the best pure scaling story in the field.

The 2025 error-correction results are the real thing, and the backing of Google, SoftBank, and NVIDIA is not the kind of validation that comes easily.

But credibility is not the same as capital, and here QuEra is outgunned. Its rivals are raising billions and, in some cases, going public, while QuEra works with a comparatively small war chest.

For researchers and enterprises who want to explore neutral-atom quantum computing today, Aquila on Amazon Braket is a genuinely useful, low-risk place to start.

For everyone else, QuEra is best understood as a long-dated option on a specific bet: that atoms, not superconducting circuits or trapped ions, will be the substrate of the first truly useful quantum computers. The science says the bet is reasonable. The timeline and the balance sheet say be patient.

What does QuEra Computing do?

QuEra Computing builds quantum computers using neutral atoms, individual rubidium atoms held in place by lasers and manipulated to perform quantum calculations. It sells access to its 256-qubit Aquila machine through Amazon Braket and is developing larger, error-corrected systems. The company is based in Boston and was spun out of Harvard and MIT.

Who founded QuEra and who runs it now?

QuEra was founded in 2018 by a group of Harvard and MIT physicists, including Mikhail Lukin, Markus Greiner, Vladan Vuletic, and Dirk Englund, with Alex Keesling as founding CEO. Since 2024 the CEO has been Andy Ory, a serial entrepreneur who previously founded Acme Packet and sold it to Oracle. Keesling continues to lead the technology side.

How much has QuEra raised and who are its investors?

QuEra has raised roughly $250 million in total. Its largest financing was a $230 million round announced in February 2025, led by Google with SoftBank Vision Fund 2 and Valor Equity Partners, and later joined by NVIDIA's venture arm. The company has not disclosed an official valuation.

What is a neutral-atom quantum computer?

A neutral-atom quantum computer stores each qubit in a single uncharged atom, trapped and moved using focused laser beams called optical tweezers. Entanglement is created by briefly exciting atoms into large "Rydberg" states so they interact. The main advantages are that all atoms are identical, coherence times are long, and the atoms can be rearranged for flexible connectivity.

Is QuEra a public company or can I buy the stock?

QuEra is a private company, and as of mid-2026 it has not filed for an IPO, so its shares are not publicly traded. Some rivals, such as IonQ and Rigetti, are public, and Quantinuum has filed to go public. Any claim of a QuEra IPO should be treated as speculation until the company confirms it.

Who are QuEra's main competitors?

QuEra's closest competitors are other neutral-atom companies, especially Atom Computing, which is partnered with Microsoft, and France's Pasqal. It also competes with different quantum technologies, including superconducting machines from IBM and Google, trapped-ion systems from IonQ and Quantinuum, and photonic efforts like PsiQuantum.

How much does it cost to use QuEra's quantum computer?

QuEra's Aquila machine is available on Amazon Braket at $0.30 per task plus $0.01 per shot, so a typical 1,000-shot job costs about $10. Dedicated reservations are priced at $2,500 per hour. These rates are set through AWS and can change over time.

Is QuEra's quantum computer actually useful yet?

QuEra's Aquila is useful today for research into optimization and quantum physics simulation, and it has supported dozens of peer-reviewed papers. However, it cannot yet solve commercially important problems faster than classical computers, and a fault-tolerant machine that could is on QuEra's roadmap for 2028 and beyond. Its most advanced error-correction results so far are laboratory demonstrations with Harvard and MIT.