Late last year, I wrote a Forbes piece about IBM, the leader in quantum computing using superconducting circuits, and promised an update on other modalities. This is the first follow-on piece and explores neutral-atom (NA) quantum, which operates at room temperature, and scales to a high number of qubits but currently does not match superconducting for maximal performance.

Fast and noisy quantum computing is the current reality — but with breakthroughs in low-noise qubits, error correction, and hybrid AI computing, the “noisy” part could soon be less of a barrier, accelerating the path toward more useful quantum systems.

Neutral-atom quantum computing is consequently emerging as a credible path to scalable, fault-tolerant quantum systems. The approach combines large qubit arrays, flexible qubit connectivity, lower costs and a plausible roadmap to hundreds or thousands of logical qubits requiring simpler fabrication than some competing modalities.

The neutral-atom approach is no longer a science project; it is becoming an attractive alternative in advanced computing, national security, materials discovery, and hybrid AI-plus-quantum infrastructure. (We note that Infleqtion is a client of Cambrian-AI Research.)

In this approach to quantum computing, atoms are cooled with lasers to temperatures just millionths of a degree above absolute zero. But, while the atoms are cold, the surrounding system operates at room temperature, avoiding the need for bulky and expensive cryogenic infrastructure.

Laser beams called optical tweezers trap and hold each atom in place, so they can be positioned and moved with nanometer precision. Quantum information is stored in the atoms’ internal energy levels, referred to as “clock states” in cesium atoms. Because neutral atoms are isolated from their environment they maintain quantum coherence for a long time, reducing errors and increasing fidelity.

Entanglement between qubits is achieved by exciting atoms into high-energy Rydberg states . These states allow atoms to interact strongly and perform high fidelity two-qubit gate operations. Neutral atom platforms consume relatively little power—on the order of kilowatts—making them energy-efficient compared to other modalities. State measurement is carried out through fluorescence: atoms emit tiny bursts of light when illuminated, allowing detectors to read out their quantum state with high accuracy.

Why Neutral Atoms are Getting Serious Attention

The neutral atom approach provides three strategic advantages for commercial adoption: large addressable arrays, higher reliability, and a credible path forward to large-scale fault-tolerant quantum advantage.

The category is also becoming more visible because the primary scientific barriers to achieving advantage have been addressed; it is now an engineering challenge. Several leading companies now have differentiated commercialization paths. PASQAL is building a strong European footprint in industrial and HPC deployments. QuEra is closely associated with major academic breakthroughs and hyperscaler cloud access. Atom Computing is pushing logical-qubit-oriented scaling and a tight Microsoft relationship. Infleqtion, by contrast, is pursuing a broader strategy that includes quantum computing, quantum sensing, atomic clocks and software, providing revenue today to fund an aggressive development roadmap to quantum advantage.

For investors, that diversity matters because it shows the neutral-atom category is not converging on a single business model. Some players are primarily selling access to frontier compute, some are building national-lab and cloud relationships, and some are using adjacent products such as sensing and timing to create earlier revenue while the computing stack matures.

The Commercial Landscape: Four Neutral-Atom Players

The table below presents the four primary players in the neutral-atom quantum computing space. While the neutral-atom space is relatively new, the lack of deep refrigeration requirements and high-scalability has attracted considerable interest, and competition.

The strategic issue for executives is that neutral atoms are no longer only about hardware elegance. The winners will need to solve the operational physics of quantum computing at scale: processor calibration, noise modeling, error correction, and low-latency integration with classical compute. This is where the category starts to intersect directly with AI infrastructure.

Why Infleqtion Deserves Special Attention

Within this field, Infleqtion occupies a particularly interesting position. It is not just another neutral-atom hardware company; it combines quantum computing, sensing, timing, software, and GPU-linked error-correction workflows in a single commercial story. That matters because the race to useful quantum computing will not be won by qubit count alone. It will be won by the vendors that can make quantum systems calibrate faster, correct errors in real time, and integrate into existing AI and high-performance computing stacks. Infleqtion already has a revenue stream from its sensing, timing, and software businesses to help fund the ongoing development of quantum computing.

Infleqtion brings together several elements that are usually scattered across different quantum companies. Its Sqale quantum computer is based on neutral atoms, supports arrays of up to 1,600 sites, reports entangling gate fidelity of 99.73% ± 0.03%, and is tightly integrated with Superstaq, the company’s compiler and orchestration software. That gives it a meaningful hardware-plus-software foundation rather than a standalone lab system.

The company also differs because it is already operating across multiple quantum-adjacent markets. Its products and programs span quantum computing, RF sensing, atomic clocks and inertial navigation, with deployments tied to defense, national labs, and space-oriented environments. From a business standpoint, that creates a more resilient commercialization profile than a pure-play compute vendor that must wait for large-scale error-corrected quantum advantage before generating meaningful revenue and profit.

Infleqtion’s public-market status also makes it easier for investors to track its strategic messaging and ecosystem posture. The company has been explicit that hybrid quantum-classical infrastructure, especially with NVIDIA, is central to its path to utility and fault tolerance.

Why NVIDIA Ising Changes The Conversation

NVIDIA launched the Ising family of open AI models for two foundational quantum workloads: calibration and error-correction decoding. This is important because these are not peripheral software tasks. They are two of the main engineering bottlenecks that determine whether a quantum system can operate reliably and scale economically.

Calibration: why it matters

Calibration is the ongoing process of measuring the behavior of a quantum processor and tuning control parameters, so the hardware stays within operating tolerances. In practical terms, calibration determines whether the machine is performing at its best possible physical error rate before a customer ever runs a useful workload. If calibration is slow, manual, or inconsistent, the result is lower uptime, lower fidelity, and a less investable platform.

NVIDIA’s Ising Calibration model is a 35-billion-parameter vision-language model designed to interpret experimental outputs and guide agentic calibration workflows. NVIDIA says the model was trained on data spanning multiple qubit modalities, including neutral atoms, and benchmarked through QCalEval, which it describes as the first benchmark for agentic quantum-computer calibration. For business readers, the key point is straightforward: better calibration means faster iteration, better hardware utilization, and a shorter path from research hardware to a repeatable commercial service.

Decoding: why it matters even more

Decoding is the classical compute task at the heart of quantum error correction. Every round of quantum error correction generates syndrome data that must be interpreted fast enough to identify likely errors and feed corrections back before noise accumulates. If the decoder is too slow, the quantum processor may improve in theory while still failing in practice because the classical side cannot keep pace.

NVIDIA says its Ising Decoding models can improve both speed and accuracy relative to conventional baselines such as PyMatching, with reported performance including up to 2.5x faster decoding and up to 3x better logical error rate performance in some conditions. NVIDIA also describes a real-time stack using CUDA-Q QEC, CUDAQ-Realtime, and NVQLink to reach microsecond-scale decoding loops, including 2.33 microseconds per round in one published configuration. For business leaders, decoding is critical because it effectively determines whether future fault-tolerant quantum systems can be run as usable accelerators rather than fragile research instruments.

Infleqtion’s Unique Position in Ising Calibration and Decoding

Infleqtion is noteworthy because it has publicly stated support for both NVIDIA Ising Calibration and NVIDIA Ising Decoding, while also showing a modality-specific path for integrating decoding into a neutral-atom system. That combination matters. Many quantum companies can say they are interested in AI for control; far fewer have publicly tied themselves to both calibration and decoding workflows inside NVIDIA’s emerging quantum-AI stack. In fact, Infleqtion is the only neutral atom company highlighted in the Ising announcements.

The calibration side matters because neutral-atom systems depend on precise control of lasers, traps, state preparation, and measurement, all of which can drift and create operational variability. Support for Ising Calibration means Infleqtion can potentially automate more of that tuning process, cut engineering labor, improve consistency, and reduce the time between maintenance, setup, and productive computation.

The decoding side may be even more strategically important because Infleqtion is not merely adopting a generic decoder. It is integrating NVIDIA Ising Decoding into a leakage-aware framework tailored to the physical realities of neutral atoms. NVIDIA’s generic decoding models were developed for scalable QEC workflows, but Infleqtion’s neutral-atom systems face additional challenges because the hardware behaves as multi-level systems, or qudits, rather than idealized two-level qubits. In those systems, atoms can leak into non-computational states or be lost entirely, complicating the syndrome patterns that a decoder must interpret.

Infleqtion says it has incorporated the Ising workflow into its Leaky simulation framework so that decoding is tested against realistic leakage and atom-loss behavior rather than against simplified textbook noise. This is strategically significant because a decoder that works only on idealized noise models may not translate into commercial performance on real hardware. In other words, Infleqtion is trying to connect the AI model not just to quantum error correction in theory, but to the actual physics of its own platform.

Why calibration and decoding are so important for investors

From an investor’s perspective, calibration and decoding are important because they are leverage points for the entire quantum P&L. Better calibration can improve uptime, shrink operating expense, reduce engineering overhead, and increase the amount of useful compute delivered per machine-hour. Better decoding can increase the effective value of each physical qubit by making logical qubits more practical, which in turn improves the economics of scaling toward useful applications.

This is the deeper significance of Infleqtion’s positioning. The company is not only selling a neutral-atom narrative; it is aligning itself with the emerging view that quantum advantage will require tightly coupled QPU-plus-GPU systems. In the same way that AI data centers became valuable only after software frameworks, networking, and accelerators matured together, useful quantum systems are likely to depend on a full stack that spans calibration, error correction, compilers, and classical acceleration.

For neutral atoms specifically, this could be a major advantage. The modality is attractive because it can scale to large arrays and flexible geometries, but that promise only converts into enterprise value if the control and correction layers are equally scalable. Infleqtion’s messaging suggests it understands that problem and is building around it earlier than many rivals.

Strategic Conclusions for the Neutral-atom Market

Three implications stand out for business leaders and investors.

- First, the neutral-atom market is increasingly separating into companies that merely build qubits and companies that build operating systems for useful quantum infrastructure.

- Second, AI is becoming part of the quantum stack itself, not just a customer workload running on future quantum machines.

- Third, the vendors most likely to create durable value are those that can bridge hardware physics, software, and data-center-class classical acceleration.

Infleqtion is compelling in that context because it sits at the intersection of all three. It has neutral-atom hardware, a software layer, adjacent revenue opportunities in sensing and timing, and now a clearly articulated position inside NVIDIA’s calibration-and-decoding framework.

The core takeaway: the most attractive quantum stories are shifting from “who has the most qubits” to “who can build the most operationally useful quantum system.” Today, Infleqtion has one of the strongest claims in the neutral-atom category to that broader, more business-ready narrative.

Disclosures: This article expresses the opinions of the author and is not to be taken as advice to purchase from or invest in the companies mentioned. My firm, Cambrian-AI Research, is fortunate to have many semiconductor firms as our clients, including Baya Systems BrainChip, Cadence, Cerebras Systems, D-Matrix, Flex, Groq, IBM, Intel, Micron, NVIDIA, Qualcomm, Graphcore, SImA.ai, Synopsys, Tenstorrent, Ventana Microsystems, and scores of investors. I have no investment positions in any of the companies mentioned in this article. For more information, please visit our website at https://cambrian-AI.com .