Fab-Like Qubits: IBM Adds HRL's Silicon Spin Architecture to Its Quantum Stack
IBM's July 23 HRL acquisition adds silicon spin qubits to its quantum portfolio—exchange-only qubits fabbed like classical chips, operating at 1 K, with HRL's 18-qubit CMOS demo as the proof point.
The acquisition
On July 23, 2026, IBM announced it had signed a definitive agreement to acquire HRL Laboratories, a research institution with deep expertise in quantum technologies, sensing, and novel materials. IBM framed the deal as complementary to its superconducting qubit roadmap—not a pivot away from it, but a bet that silicon spin qubits share enough fabrication DNA with classical chips to matter at scale.
What spin qubits are
In IBM's explainer published the same day, spin qubits encode quantum information in the intrinsic spin of electrons—a fundamental two-level property ("up" and "down") manipulated through quantum dots that confine single electrons on chip.
HRL's architecture uses exchange-only qubits: each logical qubit corresponds to three electrons in three quantum dots built from SiGe layers. Gates are applied as electrical pulses that push electrons closer or farther apart.
The practical pitch:
- Manufacturable with existing semiconductor fab tools
- High coherence times and relatively low error rates
- Operates at 1 K, versus roughly 0.015 K for superconducting architectures
What HRL has demonstrated
IBM cites HRL's recent demonstration of a digitally controlled silicon-spin quantum computer with 54 quantum dots on three rails—up to 18 qubits—running one- and two-qubit gates and small-scale error-detecting codes on a cryogenic control board manufactured in CMOS.
That is still laboratory scale. The strategic logic is long-term: spin and superconducting qubits both leverage silicon fabrication; IBM wants both modalities in its portfolio.
Beyond computation
IBM also highlighted HRL's work in quantum sensing and quantum networking—areas where IBM's historical focus has been computational. Ultra-precise sensors and novel quantum materials could feed back into more robust qubits and better error characterization.
Jay Gambetta, IBM's Director of Research, said in the blog post that HRL's team "will help IBM push even farther forward toward the frontiers of quantum innovation" and that the portfolio strengthens long-term plans to deliver useful quantum computing.
Context: trust, not just qubit count
The spin-qubit announcement arrived one week before IBM's July 30 trio of quantum-advantage papers emphasizing trusted computation—validation frameworks when classical verification fails. IBM's near-term messaging is converging: scaling modalities (superconducting and spin) while building credibility mechanisms (doped Clifford sampling, process validation, error mitigation) so beyond-classical results can be defended.
Spin qubits do not resolve verification on their own. They extend IBM's manufacturing optionality as the field argues about which fabrication path reaches fault tolerance first.
What to watch
- Regulatory closure of the HRL acquisition and integration timeline
- Whether spin-qubit demos migrate from HRL's 18-qubit device to IBM Quantum Cloud access
- How spin coherence and control compare to IBM Heron superconducting systems on identical validation benchmarks
Sources
- IBM Quantum — What are spin qubits? (July 23, 2026)
- IBM — IBM to Acquire HRL Laboratories to Advance Quantum and Advanced Computing (July 23, 2026)