# Is a Midwest Superconducting Foundry Model Viable at Commercial Scale?

SkyWater Technology and Qolab have announced a multi-year partnership to manufacture superconducting quantum devices through a coordinated Midwest corridor — SkyWater's Minnesota fabrication facility paired with Qolab's Wisconsin design and testing operation. The deal's central technical asset is SkyWater's newly launched **SC250 platform**, which Qolab will use to advance its **Quantum System-in-Package (QSiP)** architecture. QSiP integrates cryogenic wiring, low-pass filters, and amplifiers into a single compact package — targeting the cost and complexity bottleneck that currently limits wider deployment of superconducting hardware. SkyWater is backing the collaboration with dedicated quantum tool upgrades, signaling a capital commitment to quantum-specific manufacturing within existing facilities. The stated goal: transition superconducting device fabrication from bespoke custom runs to repeatable, commercially scalable wafer manufacturing under an open-access model.

The deal puts John Martinis — the former Google Quantum AI architect who now serves as Qolab's CTO — at the center of an effort to industrialize the cryogenic control stack, which has remained a persistent cost and scaling obstacle across the superconducting qubit field.

---

## The SC250 Platform and QSiP Architecture

SkyWater's SC250 is the manufacturing foundation this partnership is built on. The source material does not specify the process node, qubit count targets, or [gate fidelity](https://quantumintel.tech/glossary/gate-fidelity) specifications for the platform — claims you'll find in coverage elsewhere should be treated with skepticism until SkyWater publishes process design kits or peer-reviewed device characterization.

What the announcement does specify is functional intent. QSiP consolidates components that are currently sourced, assembled, and tested separately — cryogenic wiring, low-pass filters, and amplifiers — into a unified package. This is not a novel concept in principle: the cryogenic control stack has long been identified as a scaling chokepoint because each additional qubit demands more physical wiring routed through the limited thermal stages of a [dilution refrigerator](https://quantumintel.tech/glossary/dilution-refrigerator). Every major superconducting program, from [IBM Quantum](https://quantumintel.tech/companies/ibm) to [Rigetti Computing](https://quantumintel.tech/companies/rigetti-computing), has grappled with this wiring density problem.

Qolab's argument, as stated by Martinis, is that integrating these components at the package level can "materially improve quantum system economics." That is a commercially specific claim — and a credible one directionally — but the industry needs device-level data on [coherence time](https://quantumintel.tech/glossary/coherence-time) impact and signal integrity before QSiP's integration trade-offs can be properly assessed.

---

## From Custom Development to Repeatable Wafer Manufacturing

The framing of this partnership deserves scrutiny. The phrase "open-access and capital-efficient model" is doing a lot of work. SkyWater has previously offered foundry services to government and defense quantum programs, so the infrastructure is not being built from scratch. What this agreement specifically claims to establish is a commercial model that moves beyond one-off custom fabrication toward repeatable wafer services — giving Qolab production cost visibility and iteration speed that custom runs cannot provide.

This is a meaningful operational distinction. Custom quantum fabrication runs are slow, expensive, and highly variable. Transitioning to a wafer service model with defined process flows allows smaller quantum hardware companies to iterate designs without bearing the full overhead of a dedicated fab. The analogy is imperfect but instructive: this is roughly what GlobalFoundries or TSMC provide to ASIC designers, adapted for a process technology with far lower volume and far stricter materials and environmental requirements.

Thomas Sonderman, SkyWater's President and CEO, framed it plainly: "Quantum technologies are moving beyond the laboratory and require commercial manufacturing models adapted for their unique needs." That is accurate as a description of where the industry is, if not a particularly differentiated strategic claim.

---

## Why Martinis and Why Now

John Martinis's involvement gives Qolab credibility that a less-recognized CTO could not. His team at Google produced the 2019 supremacy demonstration and he brings deep process knowledge of superconducting transmon fabrication. His departure from Google and subsequent focus on the manufacturing and economics layer of quantum hardware — rather than raw qubit performance — is itself a signal about where the field's unsolved problems sit.

The bottleneck is no longer purely physics. Superconducting transmon qubits can be fabricated with performance adequate for near-term NISQ work and, with error correction overhead, potentially for [fault-tolerant quantum computing](https://quantumintel.tech/glossary/fault-tolerant-quantum-computing) at sufficient scale. The blocker is reproducible, high-yield manufacturing and a cryogenic control stack that doesn't require bespoke engineering for every new system. That is what SkyWater and Qolab are explicitly targeting.

---

## Industry Trajectory: Foundry Models Matter More Than Lab Records

The broader significance of this deal is not the specific companies involved — it is the pattern. The superconducting qubit field is fragmenting into distinct layers: qubit design, fabrication, cryogenic control, and software stack. Vertically integrated players like IBM retain control across most of these layers. Startups and university spinouts increasingly cannot. A commercially available, open-access superconducting foundry with defined process platforms reduces the barrier for the next tier of hardware startups to get devices fabricated without building their own fab or relying entirely on government programs like those at NIST or MIT Lincoln Laboratory.

If SkyWater's SC250 platform delivers reproducible device performance across wafers — and that is the operative "if" — it could become a meaningful reference process for academic and early-commercial superconducting programs. The Midwest geography is not incidental: Minnesota and Wisconsin sit outside the coastal quantum cluster and could anchor a domestic supply chain node distinct from the Northeast corridor.

The risk is that quantum fab processes remain sufficiently exotic that volume wafer economics don't apply at any near-term scale. Superconducting qubit yields are sensitive to substrate purity, junction uniformity, and materials interfaces in ways that CMOS foundries haven't had to solve. SkyWater is investing in tooling upgrades to address this, but the source material does not quantify what those investments are or what yield targets the partnership is designed to achieve.

---

## Key Takeaways

- **SkyWater and Qolab have announced a multi-year superconducting quantum device manufacturing partnership**, linking SkyWater's Minnesota fab with Qolab's Wisconsin design and test operation.
- **SkyWater's SC250 platform** is the manufacturing foundation; Qolab's **QSiP architecture** integrates cryogenic wiring, low-pass filters, and amplifiers into a single compact package to reduce control stack cost and complexity.
- **John Martinis**, formerly of Google Quantum AI, is Qolab's CTO — his focus on manufacturing economics rather than raw qubit records reflects where the field's real bottlenecks are.
- **SkyWater is investing in dedicated quantum tool upgrades** to support the partnership, though specific capital figures are not disclosed in the announcement.
- The deal represents a shift from custom quantum fabrication runs toward **repeatable wafer manufacturing** under an open-access commercial model — a pattern that matters for the broader ecosystem of hardware startups that cannot afford captive fabs.
- **Skeptical note:** No device performance data (coherence times, junction yield, wafer-to-wafer uniformity) has been published. The QSiP integration trade-offs for qubit performance remain uncharacterized in open literature.

---

## Frequently Asked Questions

**What is SkyWater Technology's SC250 platform?**
SC250 is SkyWater Technology's newly launched superconducting quantum device manufacturing platform, based at the company's Minnesota fabrication facility. It is the process foundation for Qolab's QSiP development, though detailed process specifications have not been publicly released as of this announcement.

**What is Qolab's Quantum System-in-Package (QSiP)?**
QSiP is Qolab's integrated packaging architecture that combines cryogenic wiring, low-pass filters, and amplifiers into a single compact package. It is designed to reduce the cost, physical size, and complexity of the cryogenic control stack — a known scaling barrier for superconducting quantum systems.

**Who is John Martinis and why does his role at Qolab matter?**
John Martinis is a leading superconducting qubit physicist who previously led the Google Quantum AI hardware team. As CTO of Qolab, his focus on manufacturing economics and system integration signals a maturation of the field beyond pure performance benchmarking toward commercial deployment infrastructure.

**How does this differ from existing quantum foundry options?**
Most accessible superconducting fabrication today runs through government-affiliated programs (NIST, MIT Lincoln Laboratory) or internal foundries at large players like IBM. SkyWater's stated open-access commercial model would offer startups and research groups a private-sector alternative with production-oriented wafer services, though the platform's device performance relative to existing options is not yet characterized in open literature.

**What does this mean for the superconducting quantum hardware supply chain?**
If SkyWater's SC250 delivers reproducible device performance at wafer scale, it establishes a domestic Midwest node for superconducting quantum fabrication outside the existing coastal cluster. This could lower barriers to entry for hardware startups and reduce dependence on government-program access for early device iterations — provided yield and performance data eventually support the commercial claims.