Quantum connections, made real
Cisco announced its new Universal Quantum Switch this week. It’s big news which seems to have gotten missed by the mainstream media. True, it is still a research prototype design (i.e. not ready for production). But this is one of these moments which people years later will be looking back at and wonder how the whole thing started.
It nicely fits into the theme of the Recurrent Patterns narrative. At least I think so.
Hard to imagine, but there were times when computers were not connected to each other. There were big computers made by vendors like IBM or DEC (anybody remember that?) and each vendor had its own idea how to connect them. There were incompatible and competing network protocols such as SNA, DECnet or XNS. Those were days.
And then came TCP/IP which was the outcome of a research project ARPANET, sponsored by DARPA.
I know, too many acronyms and boring geek talk!
I am trying to condense 4 decades of research, standardization and adoption into 2 paragraphs, but the links are a good read. They will also provide you with context about the importance of Cisco’s announcement.
Cisco has decided it is worthwhile to undertake a very expensive investment in the technology that will enable quantum computers to be able to ‘talk’ to each other.
Cisco has done this before. Its founders realized in the 1980s that there was a major gap in the effort to connect all the computers. Similar to different protocols, the communication was supported with proprietary hardware connecting them.
That’s where Cisco came in. It built a device which was able to accommodate multiple protocols. Additionally, it was the combination of hardware and software which distinguished their solution from anything on the market then. That was back in 1984.
Lots of things have changed between now and then. But here we are in 2026 and new types of computers are coming closer and closer to becoming actually useful. In the same way as 60 years ago, each vendor has a different idea and vision of how to build a quantum computer. We have Superconducting Qubits, Trapped Ion Qubits, Neutral Atom Qubits, Photonic Qubits or Topological Qubits and each is using different methods of quantum encoding.
As you can imagine, these companies are busy with building quantum computers. They have no time and perhaps even no intention to connect to the other machines. And that’s why this announcement is important.
In a few years, the same way as 60 years ago, quantum computers will come online for production use. Then, they will have to talk to each other. In order to get that far, Cisco had to design the Quantum Network Entanglement Chip and build the Quantum Networking Software Stack.
It’s not that there aren’t other companies in that space. Notable startups like Aliro, Qunnect, Nu Quantum or memQ; or heavyweights like IBM.
Yes, all this technology is at the prototype stage, but it will become more pronounced when we see peer-reviewed papers published, public demos scheduled and standards established.
And suddenly the recurrent pattern will fully emerge for everyone to see; while you already know about it.