The U.S. Just Bet $1 Billion on Quantum Chip Manufacturing

Anderon is a manufacturing service created by IBM for quantum-hardware developers. It is designed to make quantum wafers and chips without requiring every company to build its own expensive factory. The goal is to provide a stable, repeatable source for quantum hardware as the field moves toward commercial use.

Its first wafers are already moving through a production-scale, 300-millimeter line at Albany NanoTech in New York. The facility supports superconducting qubit arrays, microwave components, and through-silicon connections. Anderon plans to provide process design kits, or PDKs, so customers know how to design for its manufacturing processes.

The foundry could serve processor developers without fabs, system builders, and national laboratories. It may also help companies with small manual facilities improve production capability. Anderon will offer customization while protecting customer intellectual property and operating neutrally, including for IBM’s competitors. Other quantum technologies would require additional process development.

The U.S. Department of Commerce awarded Anderon up to $1 billion through the CHIPS and Science Act. IBM is matching that award with $1 billion in cash. IBM is also contributing intellectual property, physical assets, and skilled workers. Together, that creates a potential $2 billion commitment to the quantum manufacturing effort.

Anderon is building on IBM’s operation at Albany NanoTech, where IBM has made quantum wafers for more than five years. Its first wafers are already moving through an automated, production-scale 300-millimeter fabrication line. IBM says shifting quantum wafer research there doubled chip-development speed and increased physical complexity tenfold.

The 300-millimeter scale is intended to support more reliable and capable manufacturing. It also enables round-the-clock processing and more elaborate wiring and qubit connections. The investment is not merely for capacity. It supports process development, automation, testing, and a shared service that could eventually manufacture chips for companies beyond IBM.

The U.S. Department of Commerce is providing up to $1 billion through a CHIPS and Science Act R&D award. IBM is matching that amount with $1 billion in cash. IBM will also contribute intellectual property, assets, and skilled workers. The public-private partnership is meant to establish a dependable quantum-chip manufacturing service.

Potential customers include quantum-processor developers without fabrication plants, system builders, and national laboratories. Companies that already make chips manually could also use the service if automated production offers better capability. Rigetti Computing is already discussing the possibility with Anderon, according to its CEO, Subodh Kulkarni.

Anderon says it will protect clients’ intellectual property and operate neutrally, including for IBM’s competitors. Customers would design chips using Anderon’s process design kits, while Anderon builds them. This arrangement could let companies focus on processor designs instead of duplicating IBM’s manufacturing infrastructure and workforce.

A shared foundry separates chip design from chip manufacturing. Companies could develop quantum processors without building and operating their own large fabrication plants. That could lower duplication, provide steadier production, and let designers concentrate on improving processor architecture rather than maintaining every manufacturing step.

Anderon plans to supply process design kits, or PDKs, that define the rules for designing chips for its processes. Customers would create designs, and Anderon would fabricate them. Its automated 300-millimeter line could use repeatable patterning, deposition, and oxidation, along with in-line testing to detect process drift.

IBM reports that moving quantum wafer research to Albany doubled chip-development speed and increased physical complexity tenfold. Rigetti is interested because of capability, not simply capacity. Still, customers must show that more consistent fabrication improves qubit performance. Anderon also needs sustained demand beyond IBM to become a true industry-wide manufacturing platform.

Quantum chips are sensitive to small manufacturing differences. In superconducting qubits, the Josephson junction is a crucial component. It contains two superconductors separated by a very thin insulating barrier. The junction helps determine important electrical behavior, so its physical dimensions must be tightly controlled.

Variations in junction area or barrier thickness can change the critical current and the qubit frequency. If frequencies become too similar, unwanted overlaps can occur. Those overlaps can interfere with operation and reduce the number of working chips, lowering manufacturing yield. This makes repeatable fabrication especially important as designs become more complex.

Anderon’s automated line could reduce variation through consistent patterning, deposition, and oxidation. In-line testing could reveal process drift while wafers are being made. However, Rigetti’s Subodh Kulkarni says customers still need to measure the real improvement in qubit performance. Better manufacturing consistency matters only if it produces more useful and reliable quantum processors.

Quantum companies have several manufacturing choices besides Anderon. They can maintain their own fabrication facilities, outsource to another supplier, or use a facility focused on prototypes rather than high-volume production. The best choice depends on their designs, manufacturing needs, and ability to support a fabrication operation.

Rigetti currently makes processors on 150-millimeter wafers in a manual fab in Fremont, California. Its yield is near 50 percent, which CEO Subodh Kulkarni considers adequate for research and development while specifications change frequently. Rigetti is nevertheless discussing Anderon because automated production may offer greater capability.

Other alternatives are also developing. Lieven Vandersypen points to QuantWare’s investment in its own wafer-scale facility. Belgium’s imec supplies prototype quantum chips instead of pursuing high-volume manufacturing. These options show that Anderon will face competition. It must demonstrate useful performance gains and attract sustained demand beyond IBM.

Qubits are the basic units used to represent and process quantum information. Unlike ordinary chip elements, they must be prepared, controlled, and connected with great precision. A useful quantum computer therefore depends not only on having many qubits, but also on making their behavior predictable and coordinated.

The article focuses on superconducting qubits and their Josephson junctions. A junction’s area and insulating-barrier thickness can change critical current and qubit frequency. When frequencies overlap undesirably, qubits can interfere with one another. Repeatable patterning, deposition, oxidation, and in-line testing may help control these variations during fabrication.

The article does not provide a formal definition of qubits, so this explanation uses established quantum-computing knowledge. It does show why manufacturing matters: IBM’s automated line enabled more complex wiring and qubit connections, while customers still need evidence that consistency improves qubit performance. Better fabrication is a means, not the final measure of usefulness.