From NVIDIA AI error correction to Cisco targeting "quantum interconnection," the quantum frenzy is making huge waves again! Tech giants are competing for the commercialization of quantum computing

According to the Zhitong Finance APP, Cisco Systems (CSCO.US), one of the largest manufacturers of computer networks and internet devices in the world, showcased a high-performance switching chip aimed at the quantum computing field to global investors on Thursday, Eastern Time. Dubbed the "founder of the internet era" and riding the wave of AI infrastructure, Cisco emphasized in a statement that this switching chip will be able to connect different types of quantum computers. As the quantum craze sweeps the globe, this move represents another key step for this leader in computer hardware to push forward in the cutting-edge technology field—ultimately aiming to connect a large network composed of super quantum computing machines, just as its devices connect to existing internet systems.

Similar to other major tech companies like Alphabet's tech giant Google and the established American tech giant IBM, Cisco is also developing core technology categories closely related to quantum computers, which can leverage the properties of quantum mechanics to solve significant problems that current computer systems cannot address. However, Cisco has not joined the competition to manufacture its own quantum computers like IBM, Google, and Amazon; instead, it is actively collaborating with a range of participants to connect their machines using Cisco's exclusive technology.

Today's large quantum computers are being advanced through a combination of various technological routes—some use lasers to illuminate rubidium atoms suspended in a vacuum, while others employ superconducting technology cooled to near absolute zero.

The recent quantum computing craze is rapidly heating up in global capital markets and cutting-edge technology supply chains. For instance, "AI chip superpower" NVIDIA has launched a series of new open-source artificial intelligence models aimed at accelerating the development of quantum computing. It is reported that NVIDIA's first global quantum computing open-source AI model family, Ising, focuses on solving two key bottlenecks: quantum processor calibration and real-time decoding for quantum error correction. NVIDIA claims that Ising Calibration can compress the calibration of quantum chips from "days" to "hours," while Ising Decoding is used to accelerate the real-time decoding required for quantum error correction.

Since 2025, NVIDIA has been accelerating the integration of quantum computing into its "AI+HPC" infrastructure landscape, with quantum-classical hybrid computing stacks like CUDA Quantum / DGX Quantum, and has announced the establishment of a quantum computing laboratory in Boston, as well as a dedicated Quantum Day at GTC, emphasizing the acceleration of quantum algorithms, quantum error correction, and quantum simulation using GPU supercomputing.

Vijoy Pandey, Senior Vice President and General Manager of Cisco's Emerging Technologies and Incubation division, stated that quantum computing researchers generally believe that these technological routes for building quantum computers may each have reasonable advantages in the future. Cisco's switching chip can operate at room temperature and uses standardized telecom optical fiber cables, making it a Universal Quantum Switch capable of transitioning between different technological routes. "This means you can speak any language," Pandey said.

Jeetu Patel, President and Chief Product Officer of Cisco, stated that while large-scale quantum computing networks may not materialize until 2030, Cisco's switch chip could have the latest direct applications in the field of security features. Patel mentioned that although the chip announced on Thursday is still a prototype or a preliminary switch chip, some early typical quantum computing applications could emerge in about three years.

The Cisco switch chip does not directly address the core bottleneck of "quantum computing power," but it is laying the groundwork for quantum networking business scenarios before and after 2030. The quantum switch chip showcased by Cisco is significant not for "immediately creating a more powerful quantum computer," but for addressing a fundamental issue in the future commercialization of quantum technology: how different quantum machines can interconnect, network, and collaborate in computing. It operates at room temperature, uses standard telecom optical fibers, and attempts to act as a "translator" between different quantum systems such as rubidium atoms and superconductors, which corresponds to the infrastructure of the future quantum internet and distributed quantum data centers.

In the short term, it is most likely to first land in quantum sensing networks and security monitoring, with Cisco stating that early uses of the prototype chip could emerge in about three years; in the long term, if large-scale quantum computing is to move towards commercialization, a single-machine approach may not be sufficient. Modular interconnection, entanglement distribution, low-loss quantum networks, and cross-architecture compatibility will become critical infrastructure on par with quantum error correction.

Is the "quantum supremacy era" approaching?

The fundamental principle of quantum mechanics is that information can exist in more than one state simultaneously before being specifically measured—such as in a superposition state, as illustrated by Schrödinger's famous unfortunate cat experiment, which may be both alive and dead until the box is opened to check.

Cisco's switch chip can connect multiple quantum sensors that are currently available into a large interconnected network and put them into a so-called quantum entangled state. If a hacker—or increasingly common, a malicious AI agent controlled by hackers—appears on the network and conducts eavesdropping, the quantum sensors will be able to detect it at lightning speed, as information collection will lead to the collapse of the entangled state.

"If you can start detecting all types of behavior happening on the network through a quantum computing switch, it would almost completely change the defensive posture of countries around the world," Patel said.

Quantum computing systems leverage the properties of quantum mechanics, such as quantum superposition and quantum entanglement, to provide a completely new computing paradigm that theoretically can greatly surpass the computational power of traditional binary computers in certain specific areas. According to a statement from Google on December 9, 2024, the Willow quantum chip demonstrated astonishing performance in benchmark tests, capable of completing a "standard benchmark computation" in less than 5 minutes, while traditional supercomputers would take 10-25 years to accomplish the same task With IonQ announcing a 99.99% fidelity for dual qubit gates, and IBM deploying a classical decoder for quantum error correction on commercial AMD FPGAs achieving nanosecond-level real-time response, the industry widely predicts that key quantum inflection points such as industry-level quantum advantage and "quantum supremacy" are only three to five years away. The approaching technological critical point is transforming quantum computing from an academic topic into an urgent national security issue, especially as "practical quantum advantage" is more likely to emerge in the next 2-5 years in the form of 'narrow scenarios + hybrid computing + verifiable benefits'.

Nicolo De Masi, CEO of IonQ, a global leader in quantum computing, recently stated that significant breakthroughs and transformations in the field of quantum computing are rapidly approaching, and the so-called "quantum supremacy era" is about to arrive. As the number of qubits and gate fidelity of quantum hardware continue to improve, the concept of "quantum supremacy" mentioned by De Masi is used to mark an important threshold—where a quantum processor completes computations on a clearly defined task at a speed that classical supercomputers cannot achieve within a reasonable time.

However, during the "fault-tolerant quantum computing engineering" phase, the core bottlenecks remain quantum error correction, scaling of logical qubits, decoherence control, low-temperature/optical/control electronic system integration, and quantum-classical interconnection architecture.

The rapid development of quantum computing has already caught the attention of the capital markets, sensing investment opportunities in quantum.

A series of technological advancements and dynamics from leading quantum computing companies such as NVIDIA, Cisco, IBM, Google, Quantinuum, and Pasqal are essentially laying the hardware and software infrastructure for the potential emergence of controllable commercial quantum computing systems by 2030. In other words, quantum computing is becoming the next generation of "grand technological narrative" most sought after by global capital following the AI super wave.

NVIDIA is committed to building a "large quantum computing infrastructure layer of AI + GPU + quantum processors," attempting to connect quantum hardware manufacturers, cutting-edge research institutions, and the AI GPU supercomputing ecosystem using open-source AI models, CUDA Quantum/DGX Quantum, and even NVQLink optimized for quantum computing. NVIDIA's move is also one of the important catalysts for the recent surge in the quantum computing boom.

Recently, the Norwegian sovereign wealth fund disclosed its long positions in multiple quantum computing companies' stocks, set to buy in the fourth quarter of 2025, involving IonQ (IONQ.US), Rigetti (RGTI.US), and D-Wave Quantum (QBTS.US), the three leading U.S. stock quantum concept companies, with the broadest risk exposure to IonQ. Moreover, the collective move of quantum computing leaders like IQM and Pasqal Holding SAS, co-founded by 2022 Nobel Prize in Physics winner Alain Aspect, to go public via SPAC is a significant signal of "accelerated financing and capitalization of the quantum computing industry." **

Quantum computing — widely regarded by the quantum physics community as the "core engine of the next generation computing revolution" — is still in a relatively early developmental stage, but the acceleration of breakthroughs in this cutting-edge technology is resonating with capital enthusiasm. The "quantum computing craze" has shifted from scientific narratives to financing, public listings, valuation expansion, and a new round of technology stock narratives accelerating large-scale commercialization.

Global quantum computing has entered a milestone stage of "moving from experimental demonstration to fault-tolerant engineering," getting closer to true large-scale commercialization. The most important current progress is not merely stacking the number of physical qubits, but rather focusing on logical qubits, quantum error correction, error rate thresholds, and modular interconnectivity. For example, Google's Willow-related papers have demonstrated progress in quantum error correction below the surface code threshold, proving that as the scale of encoding increases, the logical error rate can be reduced; IBM has provided a clearer roadmap, aiming to launch Starling by 2029, achieving approximately 200 logical qubits and 100 million quantum gate operations, and subsequently advancing to larger-scale systems.

In other words, the quantum computing industry is approaching the threshold of "usable fault-tolerant quantum machines," but truly serving high-value application scenarios such as rapid drug development, material simulation, financial system optimization, and cryptanalysis — which are unattainable in the binary era — may see preliminary commercialization models in small scales as early as 2030. However, the biggest bottleneck for controllable commercialization of quantum computing still lies in error rates and scalable engineering. Qubits are extremely fragile and can be affected by decoherence, thermal noise, control errors, crosstalk, and readout errors; obtaining a reliable logical qubit often requires a large number of physical qubits for error correction coding, which brings significant costs in infrastructure hardware, control electronics, cryogenic systems, optical systems, and real-time decoding