A million-qubit quantum chip is still far off. This team just removed one big obstacle.

A team at the University of Warwick is publishing a paper this month that proposes a new way to wire up the inside of a quantum chip: let the lattice vibrations of the chip material itself carry quantum information between qubits that aren't sitting next to each other. The proposal, written by Maksym Myronov's group with collaborators at Canada's National Research Council, targets the central engineering problem that keeps today's leading quantum machines stuck below a few thousand qubits, and it does so without bolting extra microwave hardware onto the chip.

The bottleneck is geometric. In most quantum processor designs, each qubit can only share quantum information with the qubits that are physically adjacent to it. That works for small arrays, but it doesn't scale. A chip with a million qubits, the rough size most researchers treat as a useful target for error correction and practical algorithms, would be paralyzed by the wiring problem alone. Long-range coupling is the prerequisite for any path past a few thousand qubits, which is why architectures that support it (surface codes with their long bus lines, trapped-ion chains with their shared vibrational modes) get so much attention.

The Warwick proposal, published in APL Quantum on June 15, introduces what the authors call Quantum Phononic Links, or QPLs. Phonons are simply quantized vibrations in a solid, the same vibrations that carry heat and sound. In this design, phonons in compressively strained germanium grown on a silicon wafer carry quantum information between hole spin qubits, acting as a kind of internal "quantum bus." The qubit type and the bus are made of the same material, which is the architectural twist.

That's different from the two most common ways researchers currently link distant qubits. The first is microwave couplers, extra hardware that translates quantum information into a microwave signal, sends it across the chip, and decodes it on the other side. They work, but every coupler costs real estate and adds a source of noise. The second is surface acoustic waves, which use vibrations on the chip's surface. The Warwick idea pushes that vibrational approach one layer down, into the bulk of the semiconductor itself, and pairs it with hole spin qubits (quantum bits encoded in the spin of missing electrons in a germanium film grown under mechanical stress on a silicon wafer), the same lab's preferred material platform.

That material, compressively strained germanium on silicon, is the one Myronov's group has argued for years is the most promising semiconductor platform for quantum computing, in part because its natural sensitivity to vibrations makes it a clean match for a phonon-based bus. The platform is the lab's preference, not a consensus choice. Silicon spin qubits, superconducting transmons, and trapped ions each have their own scaling bets.

The paper's most quoted line is a scale claim: that information transfer is possible "in principle" across a chip up to 300 mm in diameter. A 300 mm wafer is the standard size used in commercial chip fabrication, so the number signals a manufacturing-relevant geometry rather than a small lab sample. The University of Warwick press release announcing the work is where the more ambitious language shows up. The release title and surrounding coverage describe the work as a step toward quantum computers with a million qubits. That framing belongs to the press side of the story, not the paper. The paper itself is theoretical: it lays out the mechanism, analyzes it, and identifies the 300 mm wafer as a design aspiration.

No experimental qubit-to-qubit coupling via QPLs has been reported. Coverage from The Quantum Insider and LiveScience restates the press release and adds no measured results. The paper is a clean architectural idea for an important bottleneck, with a plausible material platform behind it, and with the actual demonstration still ahead.

The work is funded by the UK Engineering and Physical Sciences Research Council under grant EP/X039757 and by NRC Canada's Quantum Sensing Program. The next concrete watch item is whether the Warwick and NRC groups can move from paper to device: fabricating hole spin qubits in compressively strained germanium on silicon, generating phonons inside the wafer, and showing that a qubit's quantum state can survive a phonon-mediated hop to a neighbor. If they can, the "quantum bus" metaphor turns from a design choice into a measured one.