The National Science Foundation just added five teams and $20 million to its quantum testbed program. Two of the new projects target quantum networking — and that split is the more interesting signal.
The U.S. National Science Foundation's National Quantum Virtual Laboratory just grew from four projects to nine. With five additional teams joining the November 2025 pilot cohort, the program is now $20 million deeper into a distributed network of quantum hardware designed to give outside researchers access to experiments they could not otherwise run.
The wire play is grant expansion: four to nine projects, $4 million each over a two-year design timeline. That framing is accurate but surface. The more useful story is what the five new teams collectively signal about where the program thinks the quantum bottleneck actually lives.
The NQVL is structured as a virtual laboratory. Instead of building one giant quantum computer, the foundation is funding a portfolio of specialized hardware, including superconducting processors, photonic links, and quantum sensors, and asking researchers to design how those pieces can be exposed to outside users as shared testbeds. The design phase is governed by the NSF 24-586 Quantum Testbeds solicitation, which sits on top of the broader NSF 23-604 NQVL opportunity launched under the National Quantum Initiative Act.
Look at the five new projects and the pattern is hard to miss. Two of the five target quantum networking, the harder unsolved problem of how to connect separate quantum machines over real distances.
The first is the Attosecond Synchronized Photonic Entanglement Network, which aims to build a roughly 60-mile distribution loop using synchronized photonic entanglement. The project description claims execution "up to 100,000x faster" than current experimental quantum networks, a number that has been carried verbatim into the aggregator coverage of the NSF announcement.
Read that claim carefully. It is program-target language, the team's own stated goal, not a measured benchmark against a deployed network. The relevant comparison is not "100,000x faster than the internet" but "100,000x faster than the experimental photonic entanglement links that exist today," most of which still lose photons faster than they can route them. The 100,000x is a roadmap stretch goal: it describes where the project wants to be, not where any quantum network currently is.
The second networking bet is Quantum Photonic Integration and Deployment, which targets field-deployable photonic links for distributed quantum sensing. Where the attosecond project is about raw speed in a controlled loop, this one is about whether photonic hardware can survive outside a lab long enough to be useful.
NSF is funding both at the same time, and that is the tell. The foundation is hedging between two networking theories: a controlled high-speed entanglement loop for connecting quantum computers, and ruggedized photonic links for connecting quantum sensors in the field. Neither is a solved problem, and the program is not pretending one is.
The other three new projects are aimed at the more familiar qubit-quality frontier. Accelerating Fault-Tolerant Quantum Logic is a co-design effort targeting the overhead that currently makes running a useful quantum algorithm prohibitively expensive. Distributed-Entanglement Quantum Sensing applies the same networking primitives to biomedical imaging. And Erasure Qubits and Dynamic Circuits for Quantum Advantage (award #2547175) is a superconducting-qubit design that makes errors detectable at the hardware level, so the classical control system can correct them before they cascade.
The split is two networking projects, one sensing project that builds on networking primitives, and two projects aimed at making individual quantum processors more useful. That distribution makes the program a more even-handed portfolio than it would look if the 100,000x claim were treated as the headline.
A few caveats. The full principal-investigator list and per-project abstracts for the four projects other than the erasure-qubits award are not yet in the public artifacts, so the analysis above is built on the five project themes rather than individual PI attribution. The cohort expansion from four to nine is anchored in the NSF release and the Quantum Computing Report coverage; the underlying program language refers to "additional teams" and "expansion" rather than a specific total, and any reversal of those numbers in a future NSF clarification would change the framing.
The two-year design timeline is the watch item. A design cohort is not a deployed testbed. What the program has actually bought is permission to find out whether the hard part of quantum computing turns out to be the qubits or the network between them, and NSF has now bet, by funding two of the five new projects on the network, that it knows which answer it is rooting for.