Why Manufacturability Could Decide the Quantum Race
A good qubit is not enough
A lab can hand build a dazzling single qubit. A useful machine needs enormous numbers of them, all behaving alike, wired up, cooled, and controlled. Roadmaps toward error corrected machines call for many physical qubits per logical qubit, so the count climbs fast. That makes manufacturing the second half of the story.
What manufacturability means in plain terms
- Yield: What share of the chips or qubits you make actually work?
- Uniformity: Do qubits made side by side behave the same?
- Reproducibility: Can another team, or another wafer, repeat the result?
- Cost and speed: How fast can a design be changed and rebuilt?
- Packaging: Wiring, fiber, cooling and control electronics around the chip.
Examples from this page cluster
| Platform | Manufacturing angle reported | Open question |
|---|---|---|
| Photonic (PsiQuantum, Xanadu) | Chips made at foundries; Xanadu reports silicon nitride waveguides on 300 mm wafers | Optical loss, linking huge numbers of parts |
| Silicon spin (Diraq, Intel) | 300 mm wafers; Intel reports 95 percent yield on Tunnel Falls; imec and Diraq report over 99 percent fidelity on foundry devices | Uniform tuning, wiring, scale beyond dozens of qubits |
| Atom precision silicon (SQC) | In house process, one week design iteration (company claim) | Scaling atom placement |
| Topological (Microsoft) | Reported as a materials and fabrication achievement by one outside physicist | Whether the underlying physics is confirmed |
Why chip factories are such a big deal
The semiconductor industry has spent decades mastering repeatable atomic scale patterning. A quantum design that fits those processes inherits tools, suppliers and know how. Reports about PsiQuantum's Omega chipset, said to be made at GlobalFoundries, and about imec and Diraq working at 300 millimeter scale, show that quantum teams are explicitly aiming at this. See PsiQuantum's story and the silicon spin comparison.
The other side: platforms with different strengths
This is not a case that photonic or silicon will win. Superconducting machines are already large and use chip fabrication too. Trapped ions and neutral atoms currently show very high quality and flexible connections, with their own scaling engineering. Whichever platform wins may be the one that combines good enough physics with the most boring, repeatable manufacturing.
The hidden items on the bill
- Cryogenics: Many platforms need cold, and large scale cooling plants are major projects. See how quantum computers are cooled.
- Control electronics: Racks of equipment per qubit does not scale. Putting electronics near or on the chip is a theme across platforms.
- Testing: Verifying millions of components is itself a hard problem.
How to spot real progress
- Look for wafer scale, multi device results, not one best chip.
- Check whether results are peer reviewed or only press releases.
- Compare like with like using benchmarks.
- Treat dates as targets. See the timeline and the 2030 to 2035 outlook.
Why we are optimistic
Quantum computing is moving from a physics question toward an engineering one, and engineering is something humanity is very good at. The fact that serious teams now talk about yield, wafers and factories is itself a sign of maturity. No one has delivered a large fault tolerant machine yet, and honesty means saying so. But the direction of travel is encouraging.
A note on assets
Technology progress does not tell you anything about the price of any token. The QNT memecoin is an independent community project with no link to the companies named here, and memecoins can lose all their value. Nothing on this page is financial advice. See the risk factors page and what QNT is.
Sources and further reading
- Quantum Zeitgeist: PsiQuantum Construct and sites
- Converge Digest: Xanadu GKP chip on 300 mm wafers
- Intel newsroom: Tunnel Falls yield
- Bits&Chips: imec and Diraq
- SQC: in house manufacturing claim
- Physics World: Majorana 1 reactions
Reported as of 2026-10-09. Company roadmaps are targets and often slip. Check each company's own announcements. This page is education, not financial advice, and the QNT memecoin is independent of every company named here.
Frequently asked questions
Why does manufacturing matter for quantum computers?
Useful machines likely need huge numbers of qubits, so cost, yield and repeatability matter as much as the quality of one qubit.
Which platform is easiest to manufacture?
Nobody knows. Photonic and silicon spin aim to use existing chip factories, but each has unresolved physics and engineering.
Does manufacturing progress affect crypto prices?
We make no price predictions. The QNT memecoin is independent of these companies. This is not financial advice.
Has anyone built a large fault tolerant quantum computer?
Not yet, as of 2026-10-09.
Keep reading
- PsiQuantum: The Big Photonic Bet, Explained
What PsiQuantum is building, why it chose light and chip factories, and what is reported about its Brisbane and Chicago sites, in plain English. - Silicon Spin Qubits Explained: Quantum Computing on Chip Factory Lines
How a single electron in silicon becomes a qubit, why that fits existing chip factories, and the main challenges, in plain English. - Types of Quantum Computers: Superconducting, Ion, Photonic and More
A guide to the main ways quantum computers are built, with the strengths and trade-offs of each approach. - Where Quantum Computing Is Heading: The Industry Roadmap to 2035
What the industry, governments and researchers say about quantum from now to 2035: logical qubits, useful applications, quantum networks and post-quantum security.
All Quantum events, catalysts and roadmap guides | Back to top | Search the site
Main pages: Quantum computing explained | Quantum and crypto | Companies | Quantum news | Glossary