Author: Miao Zheng

A few years ago, quantum mechanics was often treated as a joke: when in doubt, quantum mechanics.

But now, the joke has become an IPO prospectus.

In the past few months, three quantum computing companies—Infleqtion, Xanadu, and Horizon Quantum—have successively gone public, with several more companies waiting to enter Nasdaq.

A project that once belonged only to laboratories and science fiction movies has suddenly been brought to the public market.

The question is, has quantum computing really reached the eve of commercialization explosion?

I think not.

The most interesting part of this wave of IPOs is not that it proves quantum computing has matured, but that it exposes the true state of the industry.

Although everyone calls it quantum computing, the technological routes are varied.

Furthermore, when you carefully examine the financial reports of these companies, you will find that very few general quantum computers have been sold; instead, it is the peripheral products related to quantum computing that support the operations of these quantum computing companies.

In addition, while this business is said to be in its early stages, NVIDIA has already entered the field.

As early as 2021, NVIDIA helped researchers simulate quantum circuits on classical computers using GPUs.

Then it continued to invest in several quantum computing startups. At GTC 2025, Jensen Huang even announced the establishment of the Boston Quantum Research Center, NVAQC.

However, what Huang wants to do is not the quantum computer itself, but to transform NVIDIA into the foundational entry point of the quantum computing era.

Just like in the AI era, NVIDIA is not selling models; it sells the computational power needed for training and inference.

Whether NVIDIA can replicate its success remains an unknown. However, before that, let’s understand what the current situation of quantum computing really looks like.

Technological Routes

Although everyone calls it quantum computing, the technology differs greatly. There are four mainstream routes, each backed by completely different physical principles.

Superconducting quantum computing is the fastest route to commercialization currently.

Companies like IBM, Google, and Rigetti are all on this path.

The technical principle is to build artificial quantum bits using Josephson structures. Therefore, an extremely low-temperature environment is required, reaching levels of millikelvins.

This is indeed a cold fact: the temperature required for superconducting quantum computing is colder than outer space, which is about 2.7 kelvins.

The advantage of superconducting quantum computing is that the manufacturing process is close to traditional semiconductors, making it highly scalable in terms of qubits, but it has a short coherence time and high noise.

This route has the largest financing scale, but the dependence on cooling systems keeps costs high; a dilution refrigerator can easily cost several hundred thousand dollars.

IBM's "Golden Eye" dilution refrigerator costs over $800,000, and the annual electricity cost exceeds $100,000.

Larger systems, such as Rigetti’s cooling equipment that supports 500 qubits, can cost more than $2 million. The cooling system occupies more than 90% of the total cost of a superconducting quantum computer.

Ionic trap quantum computing is another route.

Currently, IonQ and Quantinuum are working on this. They use charged ions as quantum bits and realize quantum gate operations through laser manipulation. This route has the highest fidelity in quantum gates.

It is equivalent to a giant abacus, with charged ions as the beads; each electrical manipulation is like moving a bead. High fidelity means it performs actions more accurately and has a lower error rate.

IonQ announced in October 2025 that it achieved a dual qubit gate fidelity of 99.99%, setting a world record. Quantinuum achieved over 99.9% fidelity as early as 2024. Coherence times are also the longest, ranging from 0.2 seconds to 600 seconds, far exceeding the tens of microseconds of the superconducting route.

However, the problem with ionic traps is that the number of qubits is hard to expand.

The more ions there are, the harder they are to control. Thus, you can't simply "add more ions" to increase computational power; you must use a more complex control system to manage these ions, making it easy for ionic trap quantum computing to reach a computational ceiling.

Neutral atom quantum computing has only emerged in the past two years, but it is currently the hottest area, with Infleqtion, Pasqal, and QuEra involved.

Its principle involves using optical lattices to capture arrays of neutral atoms with optical tweezers, which means using focused laser beams to hold atoms in place. Its greatest advantage is that the number of qubits can easily reach thousands, and the coherence time is relatively long.

Infleqtion has achieved an array of 1600 physical qubits, setting the current record. The entanglement fidelity reaches 99.73%, the highest among neutral atom companies.

Infleqtion went public in February 2026, and CEO Matthew Kinsella stated, "Neutral atoms are moving from scientific advancement to commercial relevance."

Lastly, there is photonic quantum computing, which is also the easiest to understand.

Xanadu, mentioned earlier, is following this route.

Its technical principle involves using photons as information carriers, with the greatest advantage being room temperature operation, eliminating the need for vacuum or cooling systems, making it inherently suitable for the integration of quantum communication and computation.

Xanadu became the first listed photonic quantum company in March 2026. Its Aurora system claims to be the first modular, networked photonic quantum computer, with real-time error correction capabilities, and plans to achieve 500 logical qubits by 2029 to 2030.

Aurora consists of four independent server racks, interconnected by fiber optics, including 12 qubits, 35 photon chips, and 13 kilometers of fiber. It operates at room temperature, with only the photon detectors requiring a low-temperature environment.

This is the natural advantage of photonic quantum computing.

However, the fidelity of gate operations in photonic quantum computing is far inferior to that of superconducting and ionic traps.

Photons do not interact with each other naturally; two photons can pass through each other without disturbance. This makes it very challenging to achieve deterministic dual qubit gates, as light experiences loss during transmission, leading to loss of information carried by the photons.

This means that achieving the same computational power in photonic quantum computers is significantly more difficult than other routes.

Which is more reliable? In terms of technological maturity, superconducting and ionic traps are the closest to commercialization, while neutral atoms and photonic quantum routes are still in a "very promising" phase.

However, the immediate issue at hand is which route has the best cost-performance ratio, which requires considering performance, cost, deployment, and other factors altogether.

The essence of this wave of IPOs is that the capital market is being forced to vote on different technological routes for the first time. Investors are no longer satisfied with the grand narrative of "quantum computing is very important"; they want to see costs and revenue.

Xanadu's stock rose 15% on its first day of trading, but fell over 10% after hours. Horizon Quantum dropped 18% after hours. Infleqtion was valued at $1.8 billion when it went public in February, peaked at $3.8 billion, but by April, its market value had dropped to around $2.374 billion.

NVIDIA's Quantum Ambitions

When it comes to computing, NVIDIA cannot be overlooked.

NVIDIA's quantum strategy is very clear; it plans to replicate the success of CUDA and transform it into CUDA-Q, which is the quantum version of CUDA.

However, before discussing this, I need to educate everyone on a concept, which is fault-tolerant quantum computing.

The quantum bits we mentioned earlier are very fragile. Temperature, vibration, electromagnetic noise, photon loss, and even one imperfect operation can cause quantum states to drift.

Fault-tolerant quantum computing adds a complete anti-fall mechanism to these blocks.

It uses many unreliable physical qubits to form a more reliable "logical qubit." Even if some physical qubits make errors, the system can detect and correct them, then continue computing.

It's like telling a piece of information to 100 people, letting these 100 people relay the message; even if some forget or miscommunicate, at least one is likely to remember it accurately.

On the hardware level, NVIDIA has created the NVQLink platform architecture. It achieves microsecond-level latency connections between GPUs and quantum processors through RDMA over Ethernet, under 4 microseconds. This latency level is crucial for quantum error correction.

For the most advanced quantum processors, the decoding window for each error correction round is just a few microseconds. NVQLink allows GPUs to complete error correction decoding within the clock cycle of the QPU, which is a necessary condition for achieving fault-tolerant quantum computing.

On the software level, NVIDIA is working on the CUDA-Q platform and the CUDA-Q QEC library, providing a unified programming interface.

Developers can write quantum and classical hybrid applications in the same environment without concerning themselves with underlying hardware differences. The recently released version 0.6 of CUDA-Q QEC in April 2026 has already achieved deep integration with NVQLink, supporting real-time GPU decoding.

On the ecosystem level, NVIDIA collaborates with more than a dozen supercomputing centers worldwide, including Japan's G-QuAT and Singapore's National Quantum Computing Center, incorporating quantum processors into existing HPC infrastructure.

Quantinuum has announced that its latest Helios QPU and all future processors will integrate with NVIDIA GPUs via NVQLink. The Helios QPU is equipped with NVIDIA's GH200 Grace Hopper as a real-time host for quantum error correction.

Quantum computing is at a turning point from "laboratory prototypes" to "requiring large-scale classical computing support." Quantum error correction, calibration, and hybrid algorithms all need strong classical computing capabilities in real time, which is precisely NVIDIA's forte.

But here is a problem: quantum computing is not AI.

The explosion of AI occurred because deep learning is a killer application on GPUs; only GPUs can handle it well, and CPUs cannot. This is why NVIDIA shines.

At least so far, quantum computing has not yet seen a killer application.

The actual applications that can make enterprises willing to pay for quantum computing time are still not very clear.

Regarding when fault-tolerant quantum computers will be released, current industry predictions suggest that it will take another 5 to 10 years. NVIDIA, betting on both physical AI and digital twins, may not have that much time and energy to invest further in quantum computing.

In September 2025, NVIDIA invested consecutively in Quantinuum, QuEra, and PsiQuantum, covering the three major routes: ionic traps, neutral atoms, and photonics. This indicates that NVIDIA is casting a wide net, but also shows that it is uncertain which route will ultimately prevail.

If the coherence time of quantum processors significantly improves, or if a new architecture that does not rely on real-time error correction emerges, then NVQLink may have been in vain.

NVIDIA is betting on "quantum computing will inevitably move towards fault tolerance, and fault tolerance will inevitably require powerful classical computing support."

This assumption seems reasonable at present but is not the only possible technological path.

AI took about 10 years to move from the laboratory to commercialization, from 2012's AlexNet to 2022's ChatGPT.

However, quantum computing is still in an earlier stage. If it needs 10 years to commercialize, can NVIDIA wait that long?

What is the truth of the industry?

If you pay attention to the quantum computing industry, you will find that very few people are buying general-purpose quantum computers. Right now, quantum computing is making money entirely through peripheral products.

This is also the most notable issue of this wave of IPOs.

The majority of quantum computing companies that can generate real income are not their most promoted general quantum computers but instead quantum sensors, quantum clocks, control chips, software stacks, and HPC integration services.

There is still no mature commercial market formed that is scalable and replicable for general quantum computers.

In more straightforward terms, the industry is using leftover income to support a long-term mainline.

Infleqtion's main revenue sources are optical atomic clocks, quantum radio frequency receivers, and inertial sensors, applied in energy, space, and other fields.

As of June 2025, Infleqtion has sold three quantum computers and hundreds of quantum sensors, with revenue of approximately $29 million over the past 12 months and a compound annual growth rate of about 80% over the last two years. Revenue for 2026 is expected to be $40 million.

The prices of quantum sensors range from tens of thousands to hundreds of thousands of dollars. Research-grade atomic clocks and gravimeters can exceed $500,000.

With the expansion of manufacturing scale, costs are expected to decrease by an order of magnitude within the next decade, similar to solid-state lidar, which used to cost several tens of thousands but is now only $2,000.

Xanadu's situation is the same, with most of its revenue coming from quantum computing peripheral products, and the income derived from its top three customers.

Additionally, almost all publicly listed quantum companies receive significant government funding.

Xanadu has received support from DARPA projects and funding from Canada's "Quantum Champion" program. Infleqtion, IonQ, and Rigetti all have contracts with the U.S. Department of Defense and the Department of Energy.

The key question is, how long can this leftover income model be sustained?

The market scale for quantum sensing is limited.

Products like atomic clocks and inertial sensors primarily target the defense, aerospace, and research fields, not a mass market capable of supporting valuations in the hundreds of billions. And even government contracts cannot grow indefinitely; the landlord doesn't have endless provisions.

Cloud services are also unlikely to scale before quantum computers achieve "quantum supremacy." After all, current quantum computers still do not match traditional computers in cost-performance ratio.

You might say that SpaceX initially funded its Mars program through launch services, and Tesla subsidized its electric vehicle development with carbon credits.

But remember, SpaceX's launch services itself is a huge market, and rocket technology is universal; the same technology used for launching satellites is applied for Mars missions. Tesla's electric vehicles, although losing money in the early stages, at least have products that can be sold to consumers, with real market demand existing.

Quantum computing is different. No matter how they sell quantum sensors, it is tough to sustain a company with a valuation of several billion dollars in the long term.

The quantum computing industry currently finds itself in a somewhat awkward position. Technologically, there is indeed progress, but there is still a long way to go before true commercialization, especially as even the entrepreneurs themselves cannot provide a clear answer.

How far this model can go depends on two factors. One is the speed of technological breakthroughs. If a certain route suddenly achieves a significant breakthrough, such as a coherence time improvement by an order of magnitude or a substantial increase in error correction efficiency, then the entire industry’s commercialization process will accelerate.

The second is the patience of the capital market. Those who dared to invest in AI a decade ago, after seeing today’s Anthropic and OpenAI, might be even more willing to invest in quantum computing.

In my view, this wave of IPO fervor, rather than marking the beginning of quantum computing commercialization, serves as a pressure test for the capital market regarding this industry. If you can wait, then invest now.

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。