In 2023, Atom Computing did something no one had done before: it built a quantum computer with more than a thousand qubits. It did so using a technology, neutral atoms, that many believe offers one of the cleanest paths to the reliable machines the world is waiting for.
Founded in 2018 by Ben Bloom and Jonathan King, and based in Berkeley and Boulder, Atom Computing set out to build quantum computers from individual atoms held in place by lasers. It is a young company that has repeatedly punched above its weight, breaking records and partnering with some of the largest names in technology, all in pursuit of a single goal: quantum computers reliable enough to be genuinely useful.
What makes Atom worth watching is not just that it reached a thousand qubits first, but the philosophy behind how it got there. From the beginning, the company has been oriented toward fault tolerance, the reliability that practical applications demand, rather than chasing flashy but fragile demonstrations.
Atom Computing encodes its qubits in neutral atoms, specifically in the nuclear spin states of atoms suspended in a vacuum and arranged by finely controlled laser beams known as optical tweezers. This approach has several deep advantages that explain why it has attracted so much interest across the field.
First, atoms of a given element are perfectly identical by nature, so there is no manufacturing variation between qubits, a problem that can plague chip-based approaches where every qubit is slightly different. Second, neutral-atom qubits can hold their quantum state for an exceptionally long time, with coherence measured in tens of seconds, far longer than many competing technologies, which gives more time to perform reliable computation.
Third, because the atoms can be physically moved, any qubit can be brought together to interact with any other, providing flexible, all-to-all connectivity that makes many algorithms and error-correction schemes easier to implement. And fourth, the approach scales gracefully. Adding more qubits largely means trapping more atoms, rather than redesigning and refabricating a chip, which is a structural advantage on the path to the very large machines that useful quantum computing requires.
Together, these properties make neutral atoms one of the most promising platforms for reaching fault tolerance. Identical qubits, long coherence, flexible connectivity, and clean scaling are exactly the ingredients a reliable, large-scale machine needs, and Atom Computing built its whole program around them.
This is why neutral atoms have become one of the most closely watched approaches in the field, and why a relatively small company pursuing them has been able to command the attention of giants.
Atom Computing first drew wide notice with its initial 100-qubit system, but its defining moment came in 2023, when it unveiled a machine holding more than 1,100 qubits in a single array. It was the first gate-based quantum computer of any kind to exceed a thousand qubits, a symbolic and practical threshold that signaled how quickly the neutral-atom approach could scale.
The achievement mattered because qubit count, while not the only thing that matters, is a real constraint on what a machine can do, especially once error correction enters the picture. Building reliable logical qubits requires many physical qubits working together, so reaching large physical-qubit counts is a prerequisite for the fault-tolerant machines everyone is chasing. Atom got there first among gate-based systems.
Crucially, the company demonstrated that its scaling approach worked as advertised, growing the number of qubits substantially without having to reinvent the underlying machine. That validated the central promise of neutral atoms, that the path to more qubits is comparatively smooth, and it set the stage for everything that followed.
Reaching that milestone as a young, relatively small company was a striking demonstration of what a focused team with the right technology can accomplish, and it established Atom Computing as a serious contender in a field crowded with much larger players.
Atom Computing's progress attracted a powerful partner in Microsoft, and together the two companies set records that pushed the whole field forward. In a notable 2024 collaboration, they created and entangled 24 logical qubits, a record at the time for the number of error-corrected qubits demonstrated together, and ran benchmarks on even more. They also incorporated real-time detection of atom loss, a practical technique for catching and handling errors as they happen.
This work is significant because logical qubits, the stable, error-corrected units built from many physical ones, are the true currency of useful quantum computing. Demonstrating a record number of them, working together, was concrete evidence that the neutral-atom approach can deliver not just many qubits but reliable ones.
The partnership with Microsoft also brought validation and a path to market. The two committed to bringing a commercial machine with a large number of physical qubits to customers, combining Atom's hardware with Microsoft's software and cloud reach. For a young hardware company, an alliance with one of the world's largest technology firms is a powerful endorsement and an accelerant.
It positioned Atom Computing not as a laboratory curiosity but as a company on a credible path to delivering fault-tolerant quantum computing to real users.
Atom Computing has continued to push on the problem that matters most: making quantum computers reliable. In a notable 2026 milestone, the company reported a full demonstration of quantum error correction using an approach known as a toric code, showing that errors decreased as the size of the encoded system grew. This is the holy grail behavior the entire field is pursuing, evidence that adding more qubits makes a machine more reliable rather than noisier.
Demonstrating that error correction improves with scale, on its own hardware, is a major validation of Atom's fault-tolerance-first philosophy. It confirms that the platform's structural advantages, identical qubits, long coherence, and flexible connectivity, translate into real progress on reliability, not just impressive qubit counts.
The company has also looked toward the future of connecting machines together, partnering on photonic interconnects that could one day link multiple processors into larger, distributed systems. That forward-looking work reflects an understanding that scaling to truly useful machines may involve networking as well as growing individual processors.
Step by step, Atom is assembling the pieces of a fault-tolerant system, and its recent results suggest the approach is delivering on its promise.
Atom Computing's progress has earned it significant backing and recognition. It was selected by the U.S. Defense Advanced Research Projects Agency for multiple stages of a major quantum benchmarking program, a competitive validation of the seriousness of its technology. It has also raised substantial funding, including a round exceeding three hundred million dollars, and attracted interest from government sources supporting domestic quantum capability.
This combination of private investment, government interest, and partnership with a major technology company gives Atom Computing the resources and credibility to pursue its ambitious roadmap. Funding is a critical constraint in quantum computing, and Atom has assembled a strong base of support relative to its size.
The validation from DARPA and from a partner like Microsoft is, in some ways, as valuable as the money. It signals that serious, discerning organizations have examined Atom's technology closely and judged it worth betting on, which carries weight with customers and future investors alike.
For a company founded only a few years ago, assembling that kind of backing while setting technical records is a remarkable trajectory, and it positions Atom well for the next phase of scaling.
Atom Computing matters because it is pursuing one of the cleanest paths to the reliable, large-scale quantum computers the world needs, and it has the record of achievement to back the bet. It broke the thousand-qubit barrier first among gate-based machines, set logical-qubit records with Microsoft, and demonstrated error correction improving with scale, all while staying focused on the fault tolerance that genuine usefulness requires.
For anyone tracking which approaches are most likely to reach practical quantum computing, neutral atoms belong near the top of the list, and Atom Computing is among the companies leading that charge. Its blend of a promising platform, record-setting execution, and powerful partnerships makes it one of the most compelling young companies in the entire field.
Atom Computing has been clear about where it is headed. Building on its record-setting systems and its partnership with Microsoft, the company has committed to delivering a commercial machine with a large number of high-quality physical qubits, aimed squarely at customers who want to begin serious work on the path to fault tolerance. Its roadmap steadily increases the number of reliable logical qubits available, the true measure of progress toward useful machines.
The neutral-atom approach gives this roadmap unusual credibility, because scaling largely means trapping more atoms rather than redesigning the hardware from scratch. That structural advantage is why a relatively young company can credibly target the large qubit counts that fault-tolerant computing requires, and why its scaling claims have repeatedly held up in practice.
Atom has paired hardware progress with the software and error-correction techniques needed to make those qubits usable, and its collaboration with Microsoft provides a route to deliver the result through a major cloud platform. The pieces of a commercial, fault-tolerant offering are coming together, and the company has been methodical about assembling them in the right order.
For a field where timelines often slip, Atom's pattern of hitting milestones, breaking a thousand qubits, setting logical-qubit records, and demonstrating error correction that improves with scale, lends real weight to its roadmap.
The promise behind all of this is the same promise that drives the whole field, but Atom is pursuing it from one of the most favorable starting points. Reliable, large-scale quantum computers could transform the simulation of molecules and materials, accelerating the discovery of new medicines, better batteries, and novel materials, because these are quantum systems that classical computers struggle to model. They could also solve optimization problems of a scale and complexity beyond today's methods.
Neutral-atom machines are particularly well suited to the error correction these applications demand, thanks to their identical qubits, long coherence, and flexible connectivity. If any platform is positioned to reach the reliability threshold where these applications become practical, neutral atoms are a leading candidate, and Atom Computing is at the forefront of that approach.
The company's forward-looking work on connecting processors together points toward an even larger future, in which many machines are linked into distributed systems capable of tackling the hardest problems. It is an ambitious vision, but one grounded in a technology whose advantages are well matched to the challenge.
For business leaders, Atom Computing is a window into one of the most promising routes to practical quantum computing. Its combination of a favorable platform, record-setting execution, and strong partnerships makes it a company whose progress is worth following closely as the field advances toward genuine usefulness.
Jason Kumpf follows the quantum industry for what it means to business. He is Head of US Revenue at Razorpay, a board advisor, angel investor, and speaker. More about Jason.