Cisco’s Universal Quantum Switch and the rise of the quantum fabric

Cisco Systems Inc.’s new Universal Quantum Switch introduced last week is a strong proof point regarding the network’s importance in scaling quantum.

For information technology leaders, the key takeaway is that quantum is shifting from isolated computing hardware to an interconnected fabric, and Cisco has been positioning itself as the core quantum interconnect for whatever qubit technologies ultimately prevail.

Why quantum needs a network

Quantum’s big promise, namely solving problems such as molecular simulation, materials discovery, portfolio optimization and large-scale scheduling, requires on the order of 10^5 to 10^6 logical qubits, far beyond what any single system will deliver this decade. Current roadmaps top out in the thousands, and possibly in the low tens of thousands, by 2030 at best.

That gap forces a fundamental architectural shift:

• Instead of betting on a single, gigantic quantum computer, the industry is converging on distributed quantum computing. That is, many smaller processors act as a single logical machine via a quantum network, much as classical computing scaled out over Ethernet and IP.
• To do that, you cannot just move classical “results” between machines; you must move the quantum state while preserving entanglement, so the processors behave as a single, aggregated system rather than independent islands.

That’s why a network built specifically for quantum is so pivotal. In classical networks, switching silicon turned point-to-point links into the Internet. In quantum, a switch that can route entangled photons without destroying their quantum properties is the missing ingredient to turn isolated experiments into a quantum network.

What Cisco announced

Cisco’s Universal Quantum Switch is a research-grade quantum fabric element designed to route entangled photons at room temperature over standard telecom fiber while preserving quantum information across multiple encoding modalities.

While there was a bucket list of attributes to this, the most notable are:

• Quantum property-preserving switching. Conventional optical switches destroy quantum information. Cisco’s design uses an internal “quantum state converter” to take whatever encoding comes in, convert to an internal format, then reconvert on exit, without collapsing the quantum state.
• Modality universality. It supports major quantum encodings (polarization, time-bin, frequency and path) and can translate among them, so a neutral-atom system could communicate with a superconducting or photonic system through the same fabric.
• Network-native characteristics. It operates at telecom wavelengths compatible with DWDM fiber, targets nano-second reconfiguration, and is designed to share expensive elements, such as entanglement sources and detectors, across many endpoints.

Cisco couples this with its earlier entanglement source chip, which can generate roughly 200 million entangled photon pairs per second at telecom wavelengths and at room temperature, plus a stack of entanglement distribution, swapping, and teleportation protocols. In fieldwork with partner Qunnect over New York metro fiber, Cisco has already demonstrated multi-kilometer entanglement swapping at rates orders of magnitude above prior lab-only experiments.

Cisco now has the “transmitters” (entanglement source), the “fabric” (quantum switch) and the early “control plane” (compiler and orchestration software) needed to turn quantum boxes into a networked platform.

Why the network is central to quantum’s future

Quantum hardware grabs headlines, but the economic value will emerge when enterprises can treat quantum capacity as another pooled resource — much as GPUs and CPUs are consumed today via cloud and high-performance networks.

The network is the enabler in three ways:

• Scaling out in addition to scaling up. By teleporting qubits over entangled links, many modest-sized machines can function as a virtual large-scale quantum computer, accelerating the arrival of “useful” quantum computing for chemistry, finance and logistics.
• Heterogeneous quantum data centers. Different modalities excel at different algorithms (for example, trapped ions, neutral atoms and superconducting qubits), so future quantum data centers are likely to combine them. A modality-agnostic switch lets you architect for a heterogeneous future now, rather than betting on a single winner.
• Quantum-enhanced classical applications. Even before million-qubit systems exist, quantum networks enable new classical services, such as coordinated decision-making across distant trading engines (“Quantum Sync”) and fiber intrusion detection via entanglement-based sensing (“Quantum Alert”). Both rely on sharing entanglement across many endpoints, something only a scalable quantum fabric can provide.

As classical infrastructure hits physical and economic limits, the ability to add “quantum links” for specific high-value functions becomes strategically important. This is precisely where a player who understands routing, synchronization and operations at scale can differentiate.

Why Cisco is well-positioned

Quantum networking is greenfield because it involves entanglement distribution rather than conventional store-and-forward. Cisco’s approach is to build quantum networks by leveraging the existing optical fabric as much as possible, hence its focus on optical telco frequencies. A classical IP network is still required for signaling and reconfiguration. The deep knowledge required in both domains plays to Cisco’s strengths:

• End-to-end quantum networking stack. Through its incubator group, Outshift, Cisco is building hardware (entanglement chip, universal switch), software (quantum compiler, orchestration, distributed error correction) and integration with post-quantum cryptography—all anchored in an architecture that assumes heterogeneous processors from multiple vendors.
• Compatibility with existing infrastructure. Room-temperature operation at telecom wavelengths means the quantum fabric can ride on existing fiber, amplifiers and much of the optical ecosystem, rather than requiring exotic cryogenic links. That dramatically lowers deployment friction for carriers and cloud providers.
• Ecosystem and field experience. Cisco is already partnering with major modality providers, including IBM Quantum (superconducting) and Atom Computing (neutral atoms), and working with operators on metro-scale testbeds. This gives it both a voice in emerging interfaces (for example, quantum NICs and compilers) and practical experience integrating quantum gear into noisy, real-world environments.

Strategically, Cisco is playing to its strengths and approaching quantum much as it did with artificial intelligence. Rather than trying to own the entire stack, it’s becoming the fabric that brings together different vendors across modalities and locations. If quantum follows the same trajectory as classical and AI, where value concentrates around platforms that pool and route specialized resources, Cisco should be in a position to ride another rising tide.

What IT leaders should do now

Most CIOs and network leaders will not deploy a quantum switch next year, but the decisions they make over the next three to five years will determine how prepared they are when quantum moves from research to revenue.

Here are a few recommendations:

1. Treat quantum as a multivendor, networked service

Assume you will consume quantum computing from multiple providers — hyperscalers, specialized quantum clouds and possibly on-premises systems — and that those resources will need to interoperate. Architect your data center and wide-area network strategy with the expectation that quantum interconnects (for example, metro-scale entanglement links) will become another class of high-value connection, much like today’s private cloud onramps. Watch how vendors such as Cisco, IBM, and Atom define quantum NICs and APIs; those will become the “Ethernet ports” of the quantum era.

2. Start with quantum-adjacent pilots

You do not need a quantum computer to gain experience with quantum networking concepts. Explore early quantum-enhanced classical applications. For example, secure fiber monitoring, ultra-precise time synchronization or coordinated decision services in financial trading, through pilots with carriers and vendors active in this space. Use those projects to build internal expertise in entanglement-based security models, operating procedures and failure modes (including denial-of-service on quantum links) without betting on a specific qubit technology.

3. Align security and networking roadmaps

Quantum cuts both ways with security. Quantum computers threaten current cryptography, but quantum networks also enable intrinsically secure communication models. Accelerate post-quantum cryptography programs for classical control and management planes; the classical signaling around a quantum network must be hardened long before large-scale quantum adversaries exist. Track how networking vendors integrate quantum-safe algorithms into routers, switches and controllers to avoid a bifurcated “quantum-secure island” bolted onto an insecure core.

4. Build a quantum-literate architecture team

Quantum networking spans physics, optics, distributed systems and security; it will not fit neatly into any current silo. Designate a small cross-functional team (network, security, cloud and data science) to own your quantum roadmap, including vendor relationships with Cisco, IBM, hyperscalers and specialized startups. It’s important to give them a mandate to develop reference architectures for “quantum-ready” data centers and metro networks, with clear assumptions about timelines.

Quantum will not replace classical infrastructure; it will augment it where the economics justify it. Cisco’s universal quantum switch signals should simplify scaling the technology and make it less of a physics experiment and more of a roadmap IT can plan against.

Zeus Kerravala is a principal analyst at ZK Research, a division of Kerravala Consulting. He wrote this article for SiliconANGLE. 

Photo: Cisco