Vodafone: Quantum Optimisation for Network Planning with ORCA Computing

The Challenge

Planning a broadband network is one of the hardest optimisation problems a telecom operator faces. Deciding where to lay optical fibre, how to connect exchanges, cabinets, and premises, and where to place mobile base stations involves choosing among an astronomical number of possible layouts. Every kilometre of cable costs money to install and maintain, so a plan that connects the same set of points with less cable translates directly into lower rollout cost.

Many of these design tasks reduce to classic combinatorial problems. The most prominent is the Steiner Tree Problem: given a set of points that must be connected (homes, street cabinets, exchanges), find the shortest possible network of links joining them, optionally adding intermediate junction points. The Steiner Tree Problem is NP-hard, which means the time required to find guaranteed-optimal solutions explodes as networks grow. Real operators fall back on heuristics that produce good but not optimal plans.

The Vodafone and ORCA Partnership

In June 2025, Vodafone announced a partnership with UK-based ORCA Computing to apply quantum technology to modelling and optimising its networks. The documented facts:

• Vodafone runs its own network optimisation algorithms on ORCA’s PT-2 Series photonic quantum system
• In joint testing, the companies solved a Steiner Tree Problem for optical fibre cable design in minutes, a class of problem Vodafone says could otherwise take hours, weeks, or even years of classical computation
• The initial focus is optical fibre cable design, with potential applications in planning upgrades and extensions for fixed and mobile broadband, reducing cable lengths, and optimising mobile base station placement
• Longer term, the companies aim to model Vodafone’s global network, which spans more than 200 destinations and includes undersea cables and satellite systems
• The agreement is structured as a letter of intent covering continued research, algorithm development, and defined test projects, with a deeper commercial relationship to be evaluated afterwards

ORCA’s systems are photonic: they process information using single photons travelling through fibre and optical components at room temperature, a different hardware approach from superconducting or trapped-ion machines. This makes them comparatively easy to install alongside conventional data centre equipment, which matters to an operator evaluating where such a system could eventually sit in its infrastructure.

The Steiner Tree Problem, Illustrated

To see why this problem is a natural quantum optimisation target, it helps to formulate a small instance. The snippet below is a simplified educational illustration of the problem class using a classical approximation algorithm. It is not Vodafone or ORCA code, and it does not represent their algorithms or results.

import networkx as nx
from networkx.algorithms.approximation import steinertree

# Toy street graph: nodes are junctions, edge weights are cable metres
G = nx.Graph()
edges = [
    ("exchange", "j1", 400), ("j1", "j2", 250), ("j2", "cab_a", 120),
    ("j1", "j3", 300), ("j3", "cab_b", 180), ("j2", "j3", 150),
    ("j3", "j4", 220), ("j4", "cab_c", 90), ("j2", "j4", 310),
]
for u, v, w in edges:
    G.add_edge(u, v, weight=w)

# Terminals that MUST be connected: the exchange and three cabinets
terminals = ["exchange", "cab_a", "cab_b", "cab_c"]

# Classical 2-approximation of the minimum Steiner tree
tree = steinertree.steiner_tree(G, terminals, weight="weight")

total = sum(G[u][v]["weight"] for u, v in tree.edges())
print(f"Edges in plan: {list(tree.edges())}")
print(f"Total cable:   {total} m")

On a graph this small, any laptop finds a good answer instantly. The difficulty is scale: a city-level fibre plan has thousands of junctions, and the number of candidate trees grows explosively. Combinatorial problems of this shape can also be recast as QUBO formulations, binary optimisation problems suited to quantum and quantum-inspired solvers, which is the general direction operators are exploring when they test problems like this on quantum hardware.

What This Does and Does Not Show

The honest reading of the announcement is that this is a serious, structured exploration rather than a deployed capability. Vodafone has stated that joint testing solved a representative fibre design problem in minutes on the photonic system, and that the partnership will proceed through defined test projects with goals and timelines. Neither company has published benchmark comparisons against state-of-the-art classical solvers, percentage improvements, or evidence of quantum advantage, and no integration into Vodafone’s live network planning has been announced.

That makes this case study a good template for how large operators are engaging with quantum computing in practice: pick a real, costly planning problem with a clean mathematical structure, run it on real quantum hardware through a staged agreement, and decide on deeper commitment based on results.

Learn more: Quantum Optimisation Concepts

Sources

• Vodafone partners with ORCA Computing to model future networks in minutes using quantum technology (Vodafone Newsroom)