IBM announced a definitive agreement to acquire HRL Laboratories, LLC, the Malibu, California research institution jointly owned by Boeing and General Motors, in a move that gives it a fully operational silicon-spin qubit program to feed into the world's first pure-play quantum wafer foundry. The deal, with financial terms undisclosed, is expected to close by the end of Q3 2026 pending regulatory approvals. What IBM is buying is not a speculative research bet: HRL published a 54-quantum-dot, 18-qubit silicon processor in April 2026 that advances the exchange-only qubit state of the art by an order of magnitude and successfully demonstrated basic error-detecting codes. IBM's Anderon foundry in Albany, New York, built on a 300mm semiconductor line that iterates device designs 30 times faster than smaller alternatives, is the infrastructure that could take that chip from a laboratory result to an industrialized quantum hardware platform.

This is IBM's first public move toward operating two fundamentally different qubit technologies under a single manufacturing roof — a milestone no quantum hardware company has yet reached.

Spin Qubits: What IBM Did Not Have Until Today

IBM has built the world's largest deployed fleet of quantum computers — more than 90 systems globally — on a single technology: superconducting qubits, as disclosed in the company's June 2026 quantum investment announcement. Those are tiny loops of superconducting metal cooled to roughly 15 millikelvin (approximately −459.46°F/−273.135°C), colder than outer space, where quantum effects govern the circuit. They are the basis for IBM's current Nighthawk r2 processor with 120 physical qubits, which in July 2026 demonstrated a 25-times improvement in qubit reset speed. They are the hardware on which IBM's 2029 Starling and mid-2030s Blue Jay fault-tolerant systems will be built.

Silicon-spin qubits are a fundamentally different approach. Instead of superconducting circuits, they trap single electrons inside nanoscale semiconductor structures called quantum dots and encode information in each electron's quantum-mechanical spin property. HRL's specific architecture — called exchange-only (EO) qubits — uses three electrons confined in three quantum dots per qubit. The two outer electrons represent the computational states; voltage pulses applied to metal gate electrodes push those electrons closer together or further apart, triggering an exchange interaction that functions as the gate operation. No oscillating magnetic fields are required — all control is electrical, as IBM's Quantum Blog explains.

That distinction matters enormously for manufacturability. Electrical-only control is what standard CMOS semiconductor processes are built for. HRL fabricates its quantum dots from silicon-germanium (SiGe) heterostructure layers, the same material class already processed in advanced semiconductor fabs. The key operating temperature for spin qubits is approximately 1 Kelvin (about −457.87°F/−272.15°C) — still extremely cold, but roughly 67 times warmer than what IBM's superconducting systems require, potentially reducing dilution refrigerator infrastructure costs as the technology scales.

The gap in qubit count between the two technologies is real and should not be understated. IBM's superconducting Nighthawk r2 runs 120 physical qubits. HRL's most advanced silicon chip holds up to 18 EO qubits — still a research-scale device. But HRL's Thaddeus Ladd, a group leader who co-authored the landmark 2023 Nature paper, put the challenge in measured terms in the EurekAlert press release accompanying that paper: "It is hard to define what the best qubit technology is, but I think the silicon exchange-only qubit is at least the best-balanced. Real challenges remain in improving error, scale, speed, uniformity, crosstalk, and other aspects, but none of these requires a miracle."

What HRL Has Actually Built

HRL's credibility in spin qubits comes from a consistent track record of published firsts. In March 2023, the lab published in Nature the first demonstration of universal control of encoded spin qubits — meaning HRL's chip could execute the full set of logic operations needed for a programmable quantum computer, not just specialized or partial gate sets. That paper, authored by 23 HRL researchers, used silicon-germanium quantum dot arrays and demonstrated that the exchange interaction could implement single-qubit gates, CNOT, CZ, and SWAP operations.

The more recent result is more significant for IBM's purposes. HRL's April 2026 arXiv preprint describes a complete quantum processing unit: a custom-designed cryogenic CMOS controller integrated with a novel high-density superconducting ribbon cable and a low-noise EO qubit device on a single chip featuring 54 exchange-coupled quantum dots arranged across three rails, configurable to host up to 18 EO qubits. Critically, the team did not just demonstrate gate operations — they validated the system against a distance-5 repetition code and a quantum error-detecting code, comparing results with detailed simulations. The authors stated the qubit performance advances the exchange-only state of the art "by an order of magnitude."

That paper also highlighted the remaining engineering constraints plainly: questions around achieving "sufficiently low noise and a scalable control and wiring solution" are still open problems. The fabrication approach was designed specifically to address the latter — the chip combines CMOS-like wafer processing with a cryogenic controller that can be manufactured using the same silicon processes as the qubits themselves.

In July 2025, HRL released spinQICK, an open-source extension to Fermilab's Quantum Instrumentation Control Kit designed to run spin-qubit experiments on low-cost, off-the-shelf Xilinx Radio Frequency System-on-Chip (RFSoC) field-programmable gate arrays, lowering the barrier for academic and industry groups to enter the field.

What a Multi-Modality Foundry Actually Means

IBM's Anderon foundry — announced in May 2026 via a Letter of Intent with the U.S. Department of Commerce, backed by a proposed $1 billion in CHIPS Act incentives matched by $1 billion in IBM cash — operates on a 300mm wafer line in Albany's NanoTech Complex. The 300mm format is the same wafer size used in leading classical chip fabs. IBM says the Anderon line iterates device designs 30 times faster than smaller 200mm alternatives, running automated production cycles 24 hours a day, 7 days a week.

The immediate plan for Anderon is to fabricate superconducting qubit wafers — the chips that will feed IBM's Kookaburra (2026), Cockatoo (2027), and Starling (2029) systems on the existing roadmap. IBM's acquisition announcement explicitly stated that the HRL deal creates "an opportunity to partner even more closely with Anderon, including potential plans to develop spin-qubit manufacturing to scale quantum manufacturing and enable faster learning cycles," according to the press release.

That sentence, from the press release, describes something with no precedent in the quantum industry: a single foundry simultaneously developing manufacturable wafer processes for two distinct qubit architectures. Every operating quantum computer today was built by a vertically integrated company that designs, fabricates, and operates its own hardware using one technology. Anderon's stated ambition is to serve multiple vendors across multiple modalities — a model closer to TSMC than to anything currently existing in quantum hardware.

The HRL acquisition supplies what that ambition requires to become real for silicon spin qubits: a cleanroom, a device architecture already demonstrating error-detecting codes, a team that has published the field's landmark universality result, and process expertise for the SiGe heterostructure platform that is specifically compatible with industrial semiconductor fabrication.

IBM's acquisition is also strategically positioned relative to IonQ's January 2026 announcement of its planned purchase of SkyWater Technology, a 200mm CMOS quantum foundry, for $1.8 billion. Both deals reflect the same recognition that fabrication capacity has become the next competitive battleground in quantum hardware — and that waiting for academic fabs will not sustain a commercial roadmap.

Boeing, GM, and the Quantum Sensing Dimension

Neither Boeing nor General Motors is walking away from the quantum technology space following the transaction. Both companies will continue to partner with IBM on quantum applications and advanced technology development after the deal closes, IBM stated in the acquisition announcement. The arrangement signals that at least two of America's largest industrial companies see meaningful near-term value in quantum applications for aerospace simulation, materials discovery, and automotive supply-chain optimization.

That near-term value is more visible in a dimension of HRL's work that IBM highlighted alongside the qubit research: quantum sensing. HRL has developed ultra-precise quantum sensors capable of detecting subtle magnetic, gravitational, and inertial phenomena with precision that outperforms classical sensors. IBM's announcement described applications in life sciences, navigation, defense, and scientific measurement.

Quantum sensing is important context for any timeline discussion. IBM's Starling fault-tolerant quantum computer target is 2029 — three years away. Quantum sensing products built on HRL's technology could reach commercial markets significantly earlier, providing IBM with a revenue and partnership stream from the acquisition that does not depend on achieving fault-tolerant quantum computing.

IBM's Path to Starling and What HRL Adds

IBM's published quantum roadmap targets IBM Quantum Starling by 2029 — a system designed to deliver 200 logical qubits running 100 million quantum operations, approximately 20,000 times more powerful than today's deployed systems. Starling will be followed in the mid-2030s by IBM Quantum Blue Jay, projected to run 1 billion quantum operations. Both systems are built around IBM's superconducting qubit platform and its quantum low-density parity-check (qLDPC) error-correction codes, which IBM says reduce the physical qubit overhead per logical qubit by up to 90% compared to surface codes.

HRL's silicon-spin program does not change those near-term roadmap targets. The silicon platform is too early to contribute physical qubits to Starling. What it does is extend IBM's visibility beyond Starling and Blue Jay to the systems that will need to follow them — systems that will require qubit counts, densities, and manufacturing volumes that no superconducting fab approach has yet demonstrated. "As IBM looks to further extend quantum computing, HRL will bring robust knowledge of silicon-based spin qubit platforms and surrounding infrastructure that could offer new insights into how to best scale quantum computers into the next decade," the company said in its announcement.

In addition to qubits and sensing, HRL brings IBM capabilities in cryogenics, control electronics, qubit interconnects, packaging, advanced materials, high-speed and high-power communications, and advanced manufacturing — a broader infrastructure base than any single acquisition in quantum hardware has previously included.

IBM committed more than $10 billion to quantum computing over the next five years as part of a June 2026 investment announcement that spans research, manufacturing, and mergers and acquisitions, as disclosed in an SEC filing. The HRL acquisition is the first M&A transaction under that commitment.

Frequently Asked Questions

What is a silicon spin qubit and how does it differ from IBM's existing superconducting qubits?

A silicon spin qubit encodes quantum information in the spin — an intrinsic quantum property — of a single electron trapped in a nanoscale semiconductor structure called a quantum dot. HRL's specific approach, called an exchange-only (EO) qubit, uses three electrons in three quantum dots per qubit; logic gates are applied by adjusting voltages to move electrons closer together or apart. This is fundamentally different from IBM's superconducting qubits, which store quantum information in the electrical states of tiny metal loops cooled to approximately 15 millikelvin (roughly −459.46°F). Silicon spin qubits operate at around 1 Kelvin (approximately −457.87°F), which is still extremely cold but about 67 times warmer than superconducting systems require. More importantly, silicon spin qubits are fabricated using the same CMOS processes used for conventional semiconductor chips, which is why their long-term manufacturing scalability is of strategic interest to a company running a 300mm quantum wafer foundry.

Why does IBM's Anderon foundry matter for this acquisition?

Anderon is a standalone company IBM is establishing in Albany, New York, backed by a proposed $1 billion in CHIPS Act incentives from the U.S. Department of Commerce and a matching $1 billion in IBM investment. It will be the world's first pure-play quantum wafer foundry, running 300mm wafer fabrication processes 24 hours a day, 7 days a week — yielding 30 times faster device iteration than 200mm alternative facilities. Today, Anderon is building processes for superconducting qubit wafers. The HRL acquisition creates the explicit pathway to also develop spin-qubit manufacturing at Anderon — making it the first facility in the world designed to industrialize two distinct qubit architectures. No company has done that before.

What has HRL Laboratories actually demonstrated in silicon spin qubits so far?

HRL published the first demonstration of universal control of encoded spin qubits in Nature in 2023 — meaning its chip could execute any quantum logic gate, not just specialized operations. In April 2026, HRL published a more advanced result: a 54-quantum-dot processor with up to 18 configurable qubits, a custom cryogenic CMOS controller, and successful demonstrations of quantum error-detecting codes. The 2026 chip advances the exchange-only qubit field by an order of magnitude in performance. HRL also released spinQICK, a free open-source tool that lets academic and industry researchers control spin-qubit experiments on affordable off-the-shelf hardware, in July 2025. The remaining engineering challenges — noise uniformity, crosstalk, long-range coupling between qubits, and scale — are acknowledged openly by HRL's own researchers.

When will IBM close the HRL acquisition and what happens to Boeing and GM's involvement?

The deal is expected to close by the end of Q3 2026, subject to customary closing conditions and regulatory approvals, per the official announcement. Financial terms were not disclosed. Once closed, HRL will become part of IBM. Both Boeing and General Motors are expected to continue partnering with IBM on quantum applications and advanced technology development following the transaction, even though they will no longer own HRL.