Land, Sea, Air and Space:IonQ’s Full-Stack Bid to Keep America’s Quantum Lead over China

 PUBLISHED IN THE WEEK OF XI JINPING’S STATE VISIT TO WASHINGTON · 24 SEPTEMBER 2026

Beijing is building its quantum stack by decree. IonQ has built America’s by acquisition. Today’s real-time error-correction result, running on one ordinary chip, brings the fault-tolerant machine at the centre of that stack within reach, on the ground and in orbit.

IonQ had the audacity to believe it could compete with China and build an integrated quantum full-stack platform poised to become a critical foundation for American national security. Then IonQ did it.

IonQ is assembling America’s quantum stack with an indefatigable, contagious patriotic energy woven into the very fabric of many of the specialized leading quantum tech companies it has acquired, putting fabrication, control, keys, clocks and orbital links on one US balance sheet — from a trusted foundry to orbit.

The Cluster and the Company · What Washington Should Do Before the Truce Expires

____________________________________________________________________________

Friends of IonQ · Quantum Technology Integration Series, Part V · Figures current to 22 September 2026

The question this paper asks. Is China building a full-stack quantum system; how; how advanced is it; is the United States positioned to compete; and is IonQ the leading American entity to compete against China across the full stack?

The short answer. Yes, China is building a full stack, and it builds it the way it built electric vehicles and reusable rockets: as a state-directed cluster in which funding, laboratory administration, carrier ownership and statute do the integrating that a single company’s balance sheet does in the United States. It is strongest in quantum communications and enabling components, competitive on physical-qubit processors, and has published no logical-qubit or fault-tolerance result. The United States leads in science and in the number of independent routes to fault tolerance. Its weakness is that no federal instrument yet buys, trusts or secures the stack as a whole. Among American companies, IonQ — under Niccolo de Masi — is running Beijing’s full-stack strategy the American way: by acquisition, in public markets and under US law, from an accredited trusted foundry to orbit. Its 22 September real-time error decoder removes a key fault-tolerance bottleneck in simulation. This paper proposes IonQ as the reference candidate for the stack-level actions in Section 10, to be tested competitively against every other qualified American entrant. It is not the uncontested leader on every computing benchmark, and the paper says where it is not.

What the Administration should consider now. Nine actions that sit outside every program in force as of this date — the first tied to the Affiliates Rule, whose Busan-truce stay ends on 9 November and snaps back on 10 November absent an extension, two more that can be ordered within 30 days, and one that carries the stack into orbit, where China is already building (Sections 9A and 10).

Written by Friends of IonQ, whose members hold a long position in IonQ. No member receives compensation or future consideration from IonQ. Every claim carries its evidence tier. Not investment advice; not a solicitation to any government. Full disclosure on the final page.

Authors’ Note

The NEWS — 23 SEPTEMBER 2026, COLLEGE PARK, MD. IonQ Demonstrates Industry’s First End-to-End Real-Time Quantum Error Decoder. IonQ reports a real-time quantum error-correction decoder that runs on a single standard, off-the-shelf CPU, removing a bottleneck that forces fault-tolerant machines to pause and wait [CO-STATED]. In the company’s benchmarks the decoder handled circuits simulating up to 408 logical qubits across 88 memory blocks and magic-state factories, executing more than 31.5 million quantum operations, with as little as 0.02% added time [CO-STATED]; the accompanying arXiv preprint (2608.25027) reports under 0.3% at a physical error rate of 10⁻⁴ across all workloads [FACT]. IonQ’s quantum research lead called it a practical path to commercial-scale fault-tolerant quantum computing [CO-STATED]. It validates a core pillar of IonQ’s Walking Cat architecture and underpins its roadmap beyond 256 physical qubits [CO-STATED]. The magnitude of this result reshapes IonQ’s trajectory as the quantum anchor for American enterprise and national security alike. It also raises the geopolitical stakes for the global competitors of the United States and its allies — above all China, whose full-stack effort has published no comparable real-time decoding result that this report could locate [ARG].
Source: ionq.com/news/ionq-demonstrates-industrys-first-end-to-end-real-time-quantum-error-decoder

Today’s announcement by IonQ is of consequential significance for the United States and for its allies. It bears directly on whether the free world keeps the lead in quantum technology — on land, where quantum computers will sit in data centres, laboratories and command posts, and in space, where the next generation of computing, sensing and secure communication is already being placed in orbit [ARG].

What IonQ announced on 22 September 2026 is a way to keep a large quantum computer correcting its own mistakes, in real time, using nothing more exotic than a single ordinary computer chip [FACT]. Quantum bits are fragile; they pick up errors constantly. A useful quantum computer therefore works like a manuscript with a proof-reader at its shoulder: the machine keeps reporting small warning signals, and a conventional computer — the decoder — has to read those signals, work out what went wrong and tell the machine how to fix it, continuously, before the next step. If the proof-reader falls behind, the whole computation has to stop and wait. Until now, the published work kept pace only on small jobs — protecting stored information or a handful of operations — and leaned on specialised hardware such as custom chips or banks of graphics processors [FACT].

IonQ’s researchers showed a proof-reader that keeps up with full-scale programs — workloads of up to 408 error-corrected (“logical”) qubits and more than a million of the hardest class of quantum operations — while running on one off-the-shelf processor of the kind found in a high-end laptop, and adding less than a third of one per cent to the running time [FACT]. The work is a detailed simulation of IonQ’s own fault-tolerant design, published for scrutiny, rather than a run on finished hardware; the next proof is the same decoder keeping pace with a live machine [FACT]; [UNDISC]. But it removes, on paper, one of the bottlenecks that has stood between today’s quantum computers and ones that can run long, useful calculations without failing [INFER].

Why that matters to America and its allies is practical. A decoder that needs one ordinary chip rather than racks of specialised hardware is cheaper, smaller and easier to power, to secure and to supply from trusted sources [INFER]. On land, that lowers the cost of placing fault-tolerant machines where they are needed — at national laboratories, allied research centres, military bases and ships. In space, where every kilogram and every watt is scarce, it matters more: trapped-ion machines already avoid the heavy millikelvin refrigeration that other designs require, and a one-chip decoder addresses the second great burden of carrying a fault-tolerant computer into orbit [FACT]; [INFER]. And because the processor is a commodity part rather than a controlled, specialised accelerator, allies can build on the same architecture without waiting on scarce hardware [INFER].

It applies directly to China’s own full-stack race. Beijing is building every layer of a national quantum stack — its own chips, control systems, refrigerators, clouds, quantum-key satellites and computing constellations in orbit — but this report located no published Chinese logical-qubit result or real-time decoder operating at this scale [FACT]. China’s strategy is to own the stack into which a fault-tolerant machine will eventually be installed; this announcement strengthens the American claim to own the machine itself, and to build it in a trusted foundry at home [ARG]. A published decoder can be studied and imitated by others, including in China; what cannot be copied quickly is the combination this report describes — the architecture, the trusted foundry, the networks, the clocks and the assets already in orbit, held together inside one American company [INFER]. That is the race: not a single breakthrough, but whose stack a breakthrough lands in [ARG].

Authors’ Note: Today’s announcement by IonQ is a milestone-in-the-making that we believe significantly changes the landscape of the quantum ecosystem. [ARG]

Contents

Authors’ Note - 2

Executive Summary - 4

How to Read This Report - 5

Report at a Glance - 5

Key Judgments - 6

1. What “Full Stack” Means - 6

2. China’s Stack, Layer by Layer - 7

3. How the Stack Is Assembled - 9

4. How Advanced Is China? - 10

5. The American Stack - 10

6. The Stack Compared, Layer by Layer -11

7. The IonQ Case, Stated Candidly -12

8. Integration Versus Plurality: Three Scenarios to 2030 - 16

9. The Allied Dimension -17

9A. The Orbital Race: Compute, Keys and Constellations -17

10. What the Administration Should Consider Immediately -19

11. Watch List - 22

12. Honest Concessions - 23

13. What Would Change This View - 23

14. Conclusion -24

Appendix A — Entity Register, Mirrored - 24

Appendix B — The Orbital Record - 26

Appendix C — Primary Sources - 28

Disclosure - 29

Executive Summary

Part One — What China Is Building

China is building a complete quantum stack, and it is doing it as a nation rather than as a company. A domestic entity now occupies every layer. Origin Quantum builds the chip, the control system, the refrigerator and the operating system, and runs a cloud used from more than 160 countries, chiefly through its predecessor Wukong-72. USTC and China Telecom run the Zuchongzhi and Tianyan line. Hyqubit, a Tsinghua spin-out, supplies trapped-ion machines of more than 100 qubits. QuantumCTek operates quantum-secure networks across 40 cities, and China flies quantum payloads in orbit [FACT]. No Chinese company owns all of it. The Hefei National Laboratory’s control of the national program, state ownership of the carriers, a Norinco-led round of about CNY 3 billion for 10,000-qubit pilot lines, a demand plan matching more than 60 quantum firms to industrial buyers, and two 2026 statutes that can hold its specialists and mobilise its firms do the integrating instead [FACT]; [INFER].

It is strongest where quantum becomes infrastructure. China’s lead is in communications, components and deployment. In computing it is level on physical qubits — Wukong-180 carries 180 — and has published no logical-qubit or fault-tolerance result [FACT].

This is an infrastructure race, and China may be ahead in it. China does not need to win the computing race to win the stack race. It is building the infrastructure a quantum capability will be installed into — on the ground and, now, in orbit [ARG].

It is moving the stack into orbit. On 20 September China launched a satellite that identifies targets on board and downlinks only the result; a week earlier Shanghai launched a plan for gigawatt-scale computing in orbit by 2030 [FACT]. No quantum processor is in orbit yet. When one arrives, it will land in whichever orbital layer already exists (Section 9A) [ARG]. China is also the only country known to have run intercontinental quantum key exchange from its own satellites with foreign partners — and in that design the satellite’s owner sits inside every key it relays (Section 9A) [FACT] for the links; [INFER] for “only” and for the trust consequence.

It does not need to win the computing race to win the stack race. A country that already runs national quantum networks, builds its own refrigerators and operates globally used quantum clouds can absorb a fault-tolerant processor into infrastructure it already owns the day one arrives [ARG]. The scale of the effort cannot be measured from outside, and the reason is structural, not accidental [UNDISC].

Part Two — The American Answer: IonQ

SAME STRATEGY, DIFFERENT SYSTEM. Beijing’s plan is to own every layer of the quantum stack — the fab, the control electronics, the keys, the clocks, the satellites and the computer — and bind them into one national system [FACT]. Niccolo de Masi is running the same plan for the United States. Since he became CEO in February 2025, nine closed acquisitions have added exactly those layers: a trusted foundry, on-chip qubit control, quantum key distribution, optical clocks and quantum sensors, radar satellites and space laser links [FACT]. The difference is the method. China binds its stack with a national laboratory, a defence conglomerate, state carriers and an exit-control decree; de Masi binds America’s with one balance sheet, shareholder accountability, antitrust review and US trusted-supplier accreditation [FACT]; [ARG]. He is doing what Beijing is doing — the American way — and he is further along in owning the whole stack than any other American company [INFER].

The United States leads the science and has more routes to fault tolerance than China. What it lacks is an American counterpart to China’s stack — one entity that holds the layers together under American control, from the foundry to orbit. IonQ is that entity [ARG].

Where Beijing needs a ministry, a national laboratory, a defence conglomerate, two state carriers and a statute to hold its stack together, the United States has one company that owns the equivalent layers and answers for them under US trusted-supplier rules [ARG]:

• Trusted fabrication, owned and operating. SkyWater, a DMEA Category 1A trusted foundry, acquired on 31 July 2026. IonQ is the only quantum computing company on the accredited list for foundry services [FACT]; [INFER]. Commerce is paying others about $1.4 billion to establish quantum foundries that do not yet operate; IonQ already owns one, and its first integrated 256-qubit chips have been fabricated there [FACT]; [CO-STATED].
• Compute, on the path to fault tolerance. Tempo in market; Superion 256 open for orders, with deliveries in 2027 [FACT]. On 22 September IonQ announced a real-time error-correction decoder that runs on a single off-the-shelf CPU and kept pace with workloads of up to 408 logical qubits and more than a million T gates, adding under 0.3% delay — a simulation of its fault-tolerant architecture, with hardware validation still to come [FACT]; [UNDISC].
• Networks, keys, timing and orbit. ID Quantique, a national key network in Romania and a quantum network on a US utility’s fibre; Vector Atomic’s GPS-free clocks under DARPA award and its quantum sensor on the X-37B; 84 Skyloom terminals on orbit, including on the SDA’s military transport layer; Capella radar under NRO contract; and an announced space-to-space and space-to-ground quantum-key network [FACT]; [CO-STATED].
• Outside the chokepoint China is racing to close. Trapped ions do not need the millikelvin dilution refrigerators on which superconducting stacks depend [FACT].
• Scale to carry it. Q2 2026 revenue of $80.1 million, up 287%; 2026 guidance of $450–460 million including SkyWater [FACT].
• A national-security bench. A founding Space Force leader on the board; a former NGA director running IonQ Federal; a former IARPA director leading quantum systems; a former DARPA quantum program manager behind the timing and sensing line; and the CEO of a Category 1A trusted foundry (Section 7.4) [FACT].
• Allied reach. Operations in the United Kingdom, Switzerland and Romania give the Alliance a ready vehicle for a trusted allied stack [FACT]; [INFER].

What this paper asks. The Administration should put IonQ at the centre of the two actions that matter most: a Trusted Quantum Stack designation (R-3), which IonQ is closest to meeting today, and the purchase of one integrated Sovereign Quantum Stack (R-4), for which IonQ is the one American company able to bid every element from owned assets [ARG]. Open competition belongs in both, as the test IonQ should be asked to pass, not as a reason to wait [ARG]. Seven further actions — above all keeping quantum inside the Affiliates Rule before it snaps back on 10 November, and extending the stack into orbit — close the exposures China is using now (Section 10).

Two stages. IonQ has completed the first — assembling the stack. The second, integrating it so the parts operate as one enterprise and produce value none could produce alone, is the test that now matters (Section 7.3) [ARG].

The bounded claim, stated plainly. Quantinuum, IBM and Google each lead on particular compute benchmarks, and this paper says so (Section 7). The case for IonQ is ownership of the whole stack — which is the race China is actually running [INFER].

How to Read This Report

Where this sits in the Series. Part III, The Quantum Full Stack, scored American and allied platforms against each other. Part IV, An IonQ Solution for the AI Force, set out what IonQ can supply to the security layer beneath artificial intelligence, and carried The Conversion Machine and the NATO supplement Quantum and the Alliance as its China half. Those China assessments explained how Beijing mobilises money, laboratories, people and law. This Part asks a narrower and more practical question: what that machinery is actually building, layer by layer, and who in the United States can match it. It does not repeat the earlier work; where it depends on it, it says so.

Evidence tiers. Every material claim is tagged. The tiers are the Series standard.

Tier	Meaning	Worked example from this report
[FACT]	Documented in a primary or near-primary source, cited in Appendix C	Origin Quantum’s Wukong-180 carries 180 computational qubits and 251 coupling qubits (company and trade-press release, May 2026).
[CO-STATED]	A company’s own statement, not independently confirmed	IonQ states it has fabricated its first fully integrated 256-qubit processors at SkyWater.
[INFER]	A reasoned conclusion drawn from facts cited alongside it	Because Origin’s cloud is open to foreign users, China’s computing stack is being used as an instrument of reach, not only of self-reliance.
[ARG]	The authors’ argument or recommendation; contestable by design	The federal government should buy one integrated quantum stack, not only a computer.
[UNDISC]	A known unknown: information not disclosed or not collected	The budget of China’s Sci-Tech Innovation 2030 quantum megaproject.
[WATCH]	An indicator to monitor, with a date or trigger	First published application of Decree 841 to a named quantum specialist.

Three reading rules. First, physical qubits are not logical qubits: every qubit count in this report is physical unless it says otherwise. Second, opacity is a reason for uncertainty, not proof of hidden capability: where China’s program cannot be measured, the paper says so and does not fill the gap with assumption. Third, the comparison in Section 6 is about ownership and control of the stack, not about performance: a company can own every layer and still trail on a benchmark, and the paper keeps the two apart.

Style. “Department of Defense” is used throughout; executive orders since September 2025, including EO 14413, use the secondary name “Department of War.” Chinese names follow the form used in the cited source.

Report at a Glance

	Scope	Where
Question	Is China building a quantum full stack, how, how advanced, and can the US — and IonQ — compete?	Executive Summary; Key Judgments
Framework	Eight layers; three tests at each: own, control, access	Section 1
China and the US	Layer-by-layer inventories, assembly, maturity, federal instruments, cell-by-cell comparison	Sections 2–6
IonQ	The move-for-move case, the national-security bench, limits, failure modes, twelve-month proof table	Section 7
Outlook, allies, orbit	Scenarios to 2030; the Alliance chokepoint; compute, keys and constellations in orbit	Sections 8–9A; Appendix B
Action	Nine immediate actions outside all current programs	Section 10
Discipline	Watch list; Honest Concessions; What Would Change This View; Conclusion	Sections 11–14
Not covered	AI, robotics and drone lanes; classified sources; investment views	—

Key Judgments

KJ-1. China is building a full quantum stack, and it is held together by the state rather than by ownership. Every layer is now populated by a domestic entity — dilution refrigerators, chip lines, control systems, an operating system, national clouds, applications, a national quantum communications network and space links [FACT]. No single Chinese company owns them all; funding administration, carrier ownership and two 2026 statutes do the integrating [INFER].

KJ-2. China’s evidenced lead is in communications and enabling components; in computing it is competitive on physical qubits and silent on fault tolerance. Wukong-180 (180 qubits) and Tianyan-287 (105 qubits) are publicly accessible superconducting systems; Hyqubit’s HYQ-B100 exceeds 100 trapped ions. None reports a logical-qubit or fault-tolerance result [FACT].

KJ-3. The scale remains unmeasurable, and that is a reason for warning posture, not a finding of hidden capability. No Chinese quantum budget line is public at any administrative level [UNDISC]. Beijing’s record in automobiles, space and semiconductors raises the risk that public data understate timing; it does not establish undisclosed deployment [INFER].

KJ-4. The American answer is plural, newly state-backed, and still a thin customer. Washington has moved from funding research to directing it by executive order, financing it and taking equity in it [FACT]. It has not yet bought, trusted or secured an integrated stack, and its flagship milestone competition has $2.5 million appropriated against $215 million planned [FACT].

KJ-5. IonQ is the American answer to China’s stack — and has completed the first of two stages. It has assembled the layers China is building: a DMEA Category 1A trusted foundry — the only quantum computing company found on the accredited list for foundry services — and compute, networking, key distribution, timing and orbit [FACT]; [INFER] for “only”. The second stage, integrating those assets so they operate as one enterprise and create value none could create alone, is under way and is the test that matters now (Section 7.3) [ARG]. It does not lead every compute benchmark; the case for it is ownership of the whole stack, tested competitively [ARG].

KJ-6. The most time-sensitive decisions sit outside every current program. The Busan-truce stay of the Affiliates Rule ends on 9 November; Entity-Listed Chinese quantum clouds are open to American users; no federal framework trusts a quantum system end to end; and no instrument reaches the reported Russian isotope channel into China [FACT]; [ARG].

KJ-7. The stack is moving into orbit, and China is building that layer first. Over six months China networked its orbital computers, loaded twenty AI models, added new nodes on 20 September 2026 and announced gigawatt-scale orbital compute for 2030, while American and European quantum-networking satellites slipped to 2027 [FACT]. The first space-based quantum accelerator will inherit whatever orbital infrastructure is already in place [ARG].

1. What “Full Stack” Means

In one line. A full stack is not a computer; it is everything from the refrigerator to the customer, and the test at each layer is who owns it, who controls it, and who merely has access to it.

The phrase “full stack” is used loosely, including by companies describing themselves. This report fixes it to eight layers, counted from the bottom.

#	Layer	What it contains
1	Enabling components	Dilution refrigerators and cryocoolers, helium-3 and enriched isotopes, ultra-stable lasers, photonics, single-photon detectors
2	Fabrication and packaging	Wafer lines for qubit chips, ion traps and cryogenic control electronics; advanced packaging
3	Processors	The quantum processing unit, in any modality
4	Control and readout	Microwave and laser control systems, readout electronics, calibration
5	Operating system and middleware	Scheduling, compilation, error mitigation and correction toolchains
6	Cloud and distribution	How users reach the machine: owned cloud, carrier cloud, hyperscaler, on-premises
7	Applications and demand	Who buys, for what, and whether the state acts as anchor customer
8	Adjacent pillars	Quantum networking and key distribution, sensing and timing, space links, post-quantum migration

Two enablers cut across every layer rather than sitting in one: capital and talent. They are treated in Section 3 because that is where China’s model differs most from America’s.

Three tests at every layer. Own: the entity holds the asset. Control: a state or parent directs the asset without owning it, through funding, administration or statute. Access: the entity buys from a market it does not control. The distinction matters because a stack that is controlled can be as coherent as one that is owned — and one that is accessed can be cut off [ARG].

2. China’s Stack, Layer by Layer

In one line. China now has a domestic entity at every layer; the superconducting line is the most complete, the trapped-ion line is a Tsinghua spin-out, and the communications line is the most deployed quantum capability anywhere in the open record.

2.1 The superconducting line: Origin, USTC and China Telecom

Origin Quantum is the anchor. Its fourth-generation Wukong-180 launched in May 2026 on a single chip of 180 computational qubits and 251 coupling qubits, with the company describing its chip, measurement and control, environmental support and operating system as fully self-developed [FACT]. Its control system, Tianji 4.0, was unveiled in 2025 supporting more than 500 qubits [FACT]. Origin also builds dilution refrigeration in-house [FACT]. The company was added to the US Entity List in May 2024 [FACT] and closed a pre-IPO round of about CNY 3 billion on 7 June 2026, led by the state defence conglomerate Norinco with a cheque of about CNY 500 million, with proceeds earmarked in part for kilo- and 10,000-qubit chip pilot lines and a 12-inch wafer line [FACT].

One discrepancy is worth recording for anyone who cites Wukong-180. Trade press reports two-qubit gate fidelity of 99.90%; Origin’s own English release states 99% [FACT]. This report uses the company’s own figure [INFER].

USTC and China Telecom form the second node. China Telecom Quantum Group’s Tianyan cloud runs Tianyan-287, a 105-qubit system on a Zuchongzhi 3.0-like processor, following Tianyan-504 in 2024 [FACT]. Tianyan-504 was built on the 504-qubit “Xiaohong” chip developed by the Chinese Academy of Sciences Center for Excellence in Quantum Information and Quantum Physics and delivered to QuantumCTek — the clearest single artefact of state-laboratory hardware passing into a listed, state-controlled company [FACT]; [INFER].

 

Figure 1. Chinese quantum processors by reported physical-qubit count. Sources: USTC; Hyqubit; Origin Quantum; China Telecom Quantum Group (arXiv 2512.10504); Quantum Computing Report. Data only as reported; no logical-qubit claims exist to plot.

2.2 The trapped-ion line: Hyqubit

Hyqubit (华翊量子) was founded in January 2022 as a spin-out of Tsinghua University’s Center for Quantum Information, with Academician Duan Luming as founder and chief scientist [FACT]. Its first commercial machine, HYQ-A37, launched in April 2023; the second, HYQ-B100, is described by the company as controlling more than 100 ion qubits, and its first unit went to the China Mobile Research Institute [FACT]. The company sells the lasers, traps and control electronics behind the machine and runs its own cloud [FACT]. Its Series A of several hundred million RMB in July 2025 was co-led by a national social security fund vehicle [FACT].

A widely circulated summary describes Hyqubit as copying IonQ’s architecture. That is not supported. Hyqubit’s published approach rests on high-dimensional ion arrays descended from the Duan group’s large two-dimensional ion crystals, which is a different engineering path from IonQ’s electronic qubit control on chip [FACT]; [INFER].

2.3 Software, cloud and demand

Origin Pilot, Origin’s operating system, was made available for public download in February 2026 with support for superconducting, trapped-ion and neutral-atom hardware [FACT]. The claim that it is the world’s first downloadable quantum operating system comes from the Anhui Quantum Computing Engineering Research Centre, and at least one guide reports that the code itself remains proprietary; it should be described as “freely downloadable”, not “open-source” [FACT]; [UNDISC].

Origin’s cloud — chiefly through its predecessor machine, Wukong-72, live since January 2024 — has logged about 50 million remote accesses from more than 160 countries — a separate count puts it at 139 countries and more than a million computing tasks by June 2026 — [FACT], with American users reported among the most active despite the Entity Listing [FACT]. That is not a fortress. It is an instrument of reach: a Chinese computing stack that recruits foreign users, benchmarks itself on their workloads and sets a default [INFER].

On the demand side, Beijing launched the “Scenario Handshake Plan” on 16 September 2026, matching more than 60 quantum companies with industrial customers across six application areas, alongside a China Telecom-controlled innovation centre building a domestic quantum software stack [FACT]. China is building the customer, not only the supplier [INFER]. The clearest single site is the 220-kilovolt Houdian Quantum Application Demonstration Substation in Hefei, run by State Grid Anhui Electric Power and commissioned in November 2024 as China’s first operating substation to use quantum measurement, communication and computing [FACT]. It sits in the core of Hefei’s planned “World Quantum Center” and supplies power to the Hefei National Laboratory, CIQTEK and iFlytek; 85 domestically developed quantum devices across 18 categories are in daily service, including fibre and “5G + quantum” encrypted links, and products from it are deployed in Ma’anshan, Xuancheng, Huaibei and Huangshan [FACT]. State media credit it with cutting measurement error by more than 500,000 kilowatt-hours a year, and State Grid Anhui puts the market for power-quantum products above CNY 10 billion [CO-STATED]. Its grid power-flow calculations were verified on real grid topology on Origin Wukong, with no reported performance advantage over conventional computers — deployment ahead of proven computing, as Section 4 finds [FACT]; [INFER]. Next on its operator’s list: an Anhui power quantum-secure communication network, a dedicated quantum computer for power-system simulation, and a higher-voltage demonstration substation [FACT].

2.4 Communications, components and space

QuantumCTek, controlled by China Telecom since January 2025, reported 2025 revenue of RMB 310 million, up 22.53%, and its first net profit, RMB 5.39 million, with quantum-secure infrastructure across 40 cities [FACT]. It has been on the US Entity List since November 2021 [FACT]. This is the best-evidenced operational quantum capability anywhere in the open record [INFER].

Cryogenics is localising. China Electronics Technology Group unveiled a “liquid-helium-free” dilution refrigerator, the XS1000, in November 2025; it still requires helium-3/helium-4 mixture for the dilution cycle [FACT]. RUSI has documented that Western export controls accelerated this localisation [FACT].

Sensing materials show the same pattern, in China’s own words. The head of State Grid Anhui’s power quantum-sensing laboratory told People’s Daily that the project’s biggest obstacle was that foreign suppliers cut off high-quality diamond sensing material; with almost no domestic high-purity diamond capability, the team built it from scratch and now fields diamond-based quantum current sensors in grid service [FACT]. Denial did not stop the program; it created a domestic supplier [INFER].

Space: China has flown quantum payloads since Micius — which ended nearly ten years in orbit and re-entered the atmosphere in January 2026 — and has written an integrated space-ground quantum network into its five-year plan [FACT].

2.5 Corrections to claims in circulation

Claim in circulation	What the record supports
Wukong runs on China Telecom’s Tianyan network	Wukong-180 runs on Origin’s own cloud; Tianyan is China Telecom’s and runs Tianyan-287 [FACT]
Origin Pilot is open-source	Freely downloadable; source availability disputed [UNDISC]
China’s quantum stack is closed to the West	Origin’s cloud serves users in 160+ countries [FACT]
Hyqubit mimics IonQ’s exact architecture	Different approach, descended from 2D ion-crystal work [FACT]
QuantumCTek was first listed in 2024	Listed November 2021; the May 2024 action added 22 further quantum entities [FACT]
The Mobilization Law names quantum	Article 8 names no technology; quantum coverage is expert commentary [FACT]

3. How the Stack Is Assembled

In one line. In the United States a balance sheet integrates a stack; in China the integration is done by the administrator of the national program, the owner of the carriers, the investor of last resort and the statute book.

The mechanism is set out in full in The Conversion Machine and Quantum and the Alliance. What this report adds is where each mechanism binds a layer.

Mechanism	What it is	Layer it binds
Program administration	The Ministry of Science and Technology issues the Sci-Tech Innovation 2030 quantum megaproject guidelines; the Hefei National Laboratory administers the competition under them — intake, review, defence, budget review and selection — while being the largest performer [FACT]	Processors, components, communications: one administrator sets priorities across all of them
Carrier ownership	China Telecom controls QuantumCTek and the Tianyan cloud; China Mobile’s research institute took the first HYQ-B100 [FACT]	Cloud, distribution and communications
State capital	Norinco-led Origin round; national social-security capital in Hyqubit; municipal funds with published 40–50% loss tolerance in Hefei [FACT]	Fabrication and pilot lines
Demand creation	Scenario Handshake Plan; state-grid and bank deployments; the Houdian quantum demonstration substation in Hefei, 85 domestic devices in grid service [FACT]	Applications
Statute	Decree 841 exit control, in force 15 September 2026; revised National Defense Mobilization Law, in force 1 October 2026, creating “new domain” mobilisation forces [FACT]	Talent and firms, across every layer
Foreign channel	CAS–Rosatom memorandum of 20 May 2026, which CAS’s own summary says covers new photonic and quantum technologies, joint research and experiments, and personnel exchange [FACT]; implementation not observed [UNDISC]	Research and, if the reported isotope flows are confirmed, components

Two cautions carry over from the earlier assessments. Decree 841 and the Mobilization Law are cross-sector authorities; no application to a named quantum specialist or firm has been published [FACT]. And the reported Rosatom supply of helium-3 and enriched isotopes to Chinese buyers rests on a single trade outlet relaying vendor statements, with the silicon isotope’s identity unresolved [UNDISC].

The counter-case, stated fairly. China’s general public budget revenue fell 1.7% in 2025, the first outright decline in decades, with a consolidated deficit near 9% of output, and municipal venture capital may be distress substitution rather than surplus [FACT]; [INFER]. A cluster that depends on loss-tolerant local money is exposed to exactly that stress.

4. How Advanced Is China?

In one line. Ahead where quantum is a network and an instrument, level on physical qubits, and unproven where it matters most — error-corrected computation.

The honest way to answer “how advanced” is by modality and by rung, not by a single score.

Modality	Highest rung evidenced	Evidence	Tier
Communications and key distribution	Civil and state deployment at national scale	QuantumCTek, 40 cities; space-ground network a named plan target	[FACT]
Superconducting computing	Engineering system, cloud-accessible	Wukong-180; Tianyan-287; Zuchongzhi 3.0 random-circuit sampling	[FACT]
Photonic computing	Quantum-advantage demonstration	Jiuzhang-4: 1,024 squeezed-state inputs, 8,176 modes, up to 3,050 photons controlled and detected (Nature, May 2026)	[FACT]
Trapped-ion computing	Commercial prototype, delivered	HYQ-B100 at China Mobile Research Institute	[FACT]
Error correction / fault tolerance	Not publicly demonstrated	No logical-qubit result located	[FACT]
Sensing (undersea, navigation)	Field trials	Drone-mounted atomic magnetometer; diamond NV magnetometer on a submersible	[FACT]
Enabling components	Industrial production, localising	Domestic dilution refrigerators; Hefei manufacture	[FACT]
Military integration	Not publicly established	No exercise, doctrine or procurement evidence	[UNDISC]

The research picture is consistent with this. The ASPI Critical Technology Tracker, which measures research impact rather than hardware, finds the United States ahead in quantum computing with China closing, and China ahead in quantum communications and post-quantum cryptography research [FACT].

What this means. China does not need to win the computing race to win the stack race. A country that already operates national quantum-secure networks, manufactures its own refrigerators and runs globally used quantum clouds can absorb a fault-tolerant processor into infrastructure it already owns the day one arrives — from wherever it arrives [ARG]. The American question is therefore not only “who builds the first fault-tolerant machine” but “into what trusted stack does it go” [ARG].

5. The American Stack

In one line. America has more independent routes to fault tolerance than China and, since June, an executive order for every layer — but it owns fewer of its components, and it still acts more like a financier than a customer.

5.1 A plurality, not a cluster

Entity	Modality	Layers owned (illustrative)	Notable 2025–26 position
IonQ	Trapped ion	Fabrication (SkyWater), processors, control, OS, networking and key distribution, timing, space links	Owns SkyWater (closed 31 July 2026); Superion 256 orders open, deliveries 2027 [FACT]; [CO-STATED]
Quantinuum	Trapped ion	Processors, software, entropy product	Traps from Honeywell, adding GlobalFoundries; ~$100m CHIPS letter of intent [FACT]
IBM	Superconducting	Processors, software, cloud; foundry funded	$1bn CHIPS letter of intent for a quantum foundry subsidiary [FACT]
Google Quantum AI	Superconducting	Processors, in-house fabrication, software	Below-threshold error-correction result on Willow (2024); declined CHIPS terms [FACT]
PsiQuantum, Atom Computing, Infleqtion, Rigetti, D-Wave	Photonic, neutral atom, superconducting, annealing	Varies	Each ~$100m CHIPS letter of intent [FACT]

The strength of this map is that five modalities are being pursued by independent balance sheets, so a dead end in one does not stall the country [INFER]. Its weakness is that almost every American entrant accesses at least two lower layers from markets it does not control — dilution refrigerators from Finland, the United Kingdom and the Netherlands, and, at many firms, control electronics from vendors outside the United States [FACT] for the refrigerators; [INFER] for control electronics.

5.2 What Washington has already done

As of 22 September 2026 the federal instruments in force are these. They matter here because Section 10 recommends only what falls outside them.

Instrument	What it does	Tier
EO 14413, Ushering in the Next Frontier of Quantum Innovation (22 June 2026)	Updates the National Quantum Strategy; creates the QC-ADDS computing effort; a national performance-assessment centre; three Defense quantum sensor projects to be fielded by 30 September 2028; five-year sensing and networking plans; supply-chain plans including prize challenges or advance market commitments; wider access to Defense-sponsored foundry resources; expansion of the FBI’s quantum counterintelligence team; a federal quantum workforce strategy; allied alignment including Pax Silica	[FACT]
EO 14412 (2026)	Federal post-quantum migration by 2030 for key establishment and 2031 for signatures	[FACT]
EO 14409 (2026)	Insider risk, trusted partners, cyber defence of critical infrastructure	[FACT]
Commerce CHIPS quantum letters of intent (21 May 2026)	$2.013bn to nine recipients, with minority equity stakes in seven companies; solicitation still open	[FACT]
DOE Quantum Genesis Q Competition (17 September 2026)	Up to $215m, milestone-based, for systems of at least 100 logical qubits; applications close 19 October; $2.5m appropriated this fiscal year	[FACT]
National Quantum Initiative	Research funding rising from $449m (FY2019) to $968m (FY2024); $625m for five DOE centres	[FACT]
Export controls	Entity List actions in 2021, 2024 and 2025 across Chinese quantum institutes, firms and cryogenic suppliers; the Affiliates Rule suspended until early November 2026 under the Busan truce	[FACT]
AI Force (announced 19 September 2026)	Structure, authority and budget not yet published	[FACT]; [UNDISC]

 

Figure 2. Commerce CHIPS quantum letters of intent. Source: NIST, 21 May 2026, as recorded in Part IV of this Series.

The pattern in the table. The United States now directs quantum by executive order, finances it through CHIPS and owns part of it through equity [FACT]. What it has not done is act as a buyer of an integrated capability. The one milestone purchase in force is for computing alone and is almost entirely unappropriated [FACT]. China’s Scenario Handshake Plan is a demand instrument; America’s instruments are overwhelmingly supply instruments [INFER].

6. The Stack Compared, Layer by Layer

In one line. China controls every layer and owns most; the US ecosystem owns the top of the stack and accesses much of the bottom; IonQ alone owns more layers than any other American company, but still accesses the cloud and several components.

#

Layer

China (cluster)

US ecosystem

IonQ alone

1

Enabling components

CONTROL — refrigerators localised (CETC, Hefei, Origin); helium-3 import-exposed; Rosatom channel unverified

ACCESS — refrigerators from allied vendors; helium-3 from North American tritium

ACCESS — but trapped ions sit outside the millikelvin dilution chokepoint

2

Fabrication and packaging

CONTROL — pilot lines and a 12-inch line funded by state capital

OWN in-house at several firms; new foundries funded, not yet operating

OWN — SkyWater, DMEA Category 1A, fabrication and packaging

3

Processors

OWN — Origin, USTC, Hyqubit

OWN — five modalities

OWN — Tempo; Superion 256 in prototype

4

Control and readout

OWN — Tianji 4.0

MIXED — in-house at some firms, imported at many

OWN — electronic qubit control on chip (CO-STATED)

5

Operating system, middleware

OWN — Origin Pilot, free download

OWN — multiple open toolchains

OWN — IonQ Quantum OS; real-time error decoder on one CPU (simulated, 408 logical qubits)

6

Cloud and distribution

CONTROL — carrier clouds (China Telecom) and Origin’s global cloud

OWN / ACCESS — hyperscalers and IBM’s own cloud

ACCESS — hyperscalers, plus direct and on-premises

7

Applications and demand

CONTROL — Scenario Handshake Plan; SOE buyers

ACCESS — thin federal demand

ACCESS — commercial and international buyers

8a

Networking and key distribution

CONTROL — national scale, 40 cities

ACCESS — pilots; NSA sceptical of QKD for national security systems

OWN — ID Quantique; EPB network; Romania deployment

8b

Sensing and timing

CONTROL — field trials

OWN — several firms; EO 14413 fielding target 2028

OWN — Vector Atomic, DARPA production award

8c

Space links

CONTROL — Jinan-1 in orbit; Micius retired Jan 2026; plan target

OWN — commercial constellations

OWN — Skyloom, 84 terminals on orbit; Capella

Table 6. Own / control / access by layer. Cells summarise facts cited in Sections 2, 5 and 7 [FACT]; the classification into own, control and access is the authors’ judgment [INFER]. This is an ownership map, not a performance ranking.

Reading the table. Three things stand out. China’s weakest layer is the lowest — helium-3 and the physics of dilution refrigeration — which is precisely where the Alliance holds leverage and where that leverage is depreciating [INFER]. America’s weakest layers are demand and components: it lacks a state customer and it imports much of the cold [INFER]. And IonQ is the only American column entry that is “own” from fabrication through space links, which is the structural argument of this paper — and also a concentration risk, discussed in Section 7 [INFER].

7. The IonQ Case, Stated Candidly

In one line. IonQ mirrors China’s cluster inside one American company; that is its strength for the stack-level actions in Section 10, and it is also why its execution risk matters to the country and not only to shareholders.

7.1 Beijing’s playbook, run by an American company

Move for move. Every layer China is building through the state, IonQ has acquired or built under Niccolo de Masi [FACT]; [INFER]. The table sets the two side by side.

Layer	China’s move (state-directed)	IonQ’s move (under de Masi)
Trusted fabrication	Norinco-led CNY 3bn round for 10,000-qubit pilot lines and a 12-inch wafer line	SkyWater acquired — a DMEA Category 1A trusted foundry, where IonQ’s first 256-qubit chips were fabricated
Qubit control	Origin’s Tianji 4.0 control system, built in-house	Oxford Ionics acquired — electronic qubit control on chip
Keys and networks	QuantumCTek under China Telecom; 40-city quantum-secure network	ID Quantique acquired; Romania national deployment; EPB quantum network
Timing and sensing	State-laboratory clocks and field-trialled quantum sensors	Vector Atomic acquired — DARPA optical clocks; quantum inertial sensor flying on the X-37B
Space links	Micius (retired January 2026), Jinan-1; laser-linked computing satellites	Skyloom acquired — 84 optical terminals on orbit
Integrated photonics	Lasers, detectors and photonics localised under export-control pressure	Nexus Photonics acquired — integrated photonics for miniaturised, mass-manufacturable quantum systems (June 2026)
Entanglement hardware	Micius-lineage entanglement sources	Qubitekk acquired — US entanglement-distribution hardware (January 2025, weeks before de Masi became CEO)
Imaging and ISR	Chaozhisuan-1: target identification in orbit	Capella acquired — radar satellites under NRO contract
Fault tolerance	No published logical-qubit or real-time decoding result located	Walking Cat architecture; real-time decoder on one CPU, announced 22 September 2026 (simulation)
Integration mechanism	National laboratory administration, state capital, carrier ownership, Decree 841, Mobilization Law	One company, one balance sheet, public-market disclosure, FTC review, DMEA accreditation
Demand	Scenario Handshake Plan: the state builds the customer	Commercial and international buyers; federal demand still thin — Washington’s gap, not IonQ’s

Table 7.1. Entries as cited in Sections 2, 3, 7, 9A and Appendix B [FACT]; the pairing is the authors’ [INFER]. Nine acquisitions have closed since de Masi became CEO in February 2025: ID Quantique (controlling stake, May 2025), Lightsynq (June 2025), Capella (July 2025), Oxford Ionics (September 2025), Vector Atomic (2 October 2025, about $181.5m in stock), Skyloom (late January 2026), Seed Innovations (30 January 2026), Nexus Photonics (30 June 2026, about $76m) and SkyWater (31 July 2026); Qubitekk closed on 6 January 2025, weeks before he became CEO [FACT] — IonQ 8-Ks, 10-Q, Form ARS and releases.

Why the parallel matters to Washington. When Beijing does this, it takes a ministry, a national laboratory, a defence conglomerate, two carriers and two statutes [FACT]. When an American CEO does it, it takes capital, execution and a customer — and the customer is the one piece the United States government has not yet supplied [ARG]. China’s pursuit of this strategy at national scale does not prove it right; it shows that Beijing considers full-stack integration strategically decisive. The question for the Administration is whether it will back the American version as deliberately as Beijing backs its own [ARG].

 

Figure 3. The stack assembled layer by layer: nine acquisition closings since Niccolo de Masi became CEO (May 2025 – July 2026), Qubitekk shown in grey as a pre-CEO closing, and the in-house real-time decoder announced on 22 September 2026 (a simulation result). Sources: IonQ releases, 8-K exhibits and Form ARS FY2025.

7.2 IonQ’s assets, layer by layer

Beyond Table 7.1, five facts complete the picture. SkyWater cost about $1.8 billion and closed on 31 July 2026 after an FTC tie vote [FACT]. IonQ reports 99.99% two-qubit gate fidelity, achieved in 2025, and its Superion 256 chips carry on-chip electronic qubit control from the $1.07 billion Oxford Ionics acquisition [FACT]; [CO-STATED]. Its real-time decoder handled simulated workloads of up to 408 logical qubits and over a million T gates on one 12-core CPU, with under 0.3% added delay at a physical error rate of 10⁻⁴ and under 12% at 5×10⁻⁴ — hardware validation pending [FACT]; [UNDISC]. Vector Atomic’s optical clocks sit under a DARPA award of $28 million, up to $58 million with option [FACT]. And the company reported Q2 2026 revenue of $80.1 million, with full-year guidance of $450–460 million including SkyWater from 31 July (organic guidance $280–290 million) [FACT].

The single sentence that carries the argument: China assembles its stack across a laboratory, a carrier, a defence conglomerate and a statute; IonQ has assembled the American equivalent inside one accountable company, under US trusted-supplier rules [ARG]. That makes IonQ the natural reference candidate for any federal action that buys, trusts or secures the stack as a whole — the actions in Section 10 [ARG].

22 September 2026: the missing layer, filled on paper. Fault tolerance needs a classical decoder that keeps pace with every error-correction cycle, or the machine stalls. Earlier work decoded a memory or a handful of operations on specialised FPGAs, GPUs or ASICs; IonQ’s team decoded full application workloads — up to 408 logical qubits and over a million T gates — end to end on one standard CPU, with under 0.3% added delay at a physical error rate of 10⁻⁴ [FACT]. That is the layer between the qubits and the answer, and IonQ now owns a published design for it that runs on commodity hardware [ARG]. It is a simulation of IonQ’s own architecture, and the company’s claim to be first is its own [CO-STATED]; the test that matters next is the same decoder keeping pace with real hardware [UNDISC].

7.3 Assembly is not integration

Stage 1 was assembling the stack. Stage 2 is making it operate as one enterprise. IonQ has done something rare: brought fabrication, control, keys, clocks, sensing, networking and orbital links under one company [FACT]. But ownership is not integration. A collection of excellent businesses becomes a system only when it runs on one set of roadmaps, architecture and decisions — and produces results none of the parts could produce alone [ARG].

What the second stage requires, and what an outside reader should look for:

Integration mechanism	What it would look like	Where to see it
Common roadmap	One sequenced technology roadmap across compute, networking, sensing and space, rather than parallel unit plans	Investor Day materials; product announcements
Shared architecture	Components designed for each other — on-chip control fabricated in the trusted foundry, the decoder built for the same fault-tolerant design, networking and repeaters built to link IonQ machines	Technical papers; product specifications
Dependency management	Explicit handling of internal dependencies — SkyWater capacity for IonQ versus merchant customers, shared photonics and laser supply	Filings; foundry disclosures
Priorities	Visible choices about what to fund, slow or stop across the portfolio	Capital allocation; programme announcements
Decision rights	Clear authority over cross-unit trade-offs, rather than federated subsidiaries	Organisation and segment structure
Customer solutions	Offers that combine units — compute with networking and timing, secure links with orbital assets — sold as one	Contracts and customer announcements
Accountability	Integration targets owned by named leaders and reported against	Earnings calls; segment reporting

Table 7.3. The authors’ integration test [ARG]. Early signs exist: IonQ’s first 256-qubit chips combine Oxford Ionics’ on-chip control with SkyWater fabrication, and the decoder was designed for the same Walking Cat architecture [FACT]; [CO-STATED]. Most of the mechanisms above are not yet visible from outside [UNDISC].

Value realization is the proof. The final test is whether the combinations produce products, customers, contracts, technical capabilities or revenue that the individual companies could not have produced on their own [ARG]. That is how a reader can tell whether integration is creating value or simply enlarging the portfolio — and it is why value realization now sits in the proof table (Section 7.7).

7.4 The national-security bench

The people are part of the stack. China’s stack is staffed by its national laboratory, its academy and its defence conglomerates. IonQ has recruited, and acquired, people who have run American intelligence, space, defence research and trusted-manufacturing institutions [FACT]; [INFER].

Name	Role at IonQ	National-security background	Where it counts in this report
General John W. “Jay” Raymond	Board of Directors, since September 2025	Former Chief of Space Operations and founding leader of the US Space Force, with nearly four decades of national-security service [FACT]; described by IonQ as the “Father of the U.S. Space Force” and by de Masi as the only four-star general since 1947 to serve as a four-star in two services [CO-STATED]	The orbital race (Section 9A); R-9
Robert Cardillo	Executive Chairman, IonQ Federal; IonQ director since 2024	Sixth Director of the National Geospatial-Intelligence Agency; national intelligence advisor to the President, driving the daily intelligence briefing; the only person to have led analytic operations at ODNI, DIA, NGA and the Joint Staff [FACT]	R-6 (measuring China); a federal stack buyer (R-4)
Dr. Rick Muller	Vice President of Quantum Systems, leading quantum computing systems development	Former Director of IARPA, running its quantum computing and AI programs; led Sandia’s Quantum and Advanced Microsystems group; directed DOE’s Quantum Systems Accelerator [FACT]	Fault tolerance and the 22 September decoder; Quantum Genesis
Dean Acosta	Chief Corporate Affairs and Government Relations Officer	Former Senior Vice President and Chief Communications Officer at Lockheed Martin; NASA press secretary through the Columbia crisis and return to flight [FACT]	Engagement with the Administration and Congress (Section 10)
Dr. Jamil Abo-Shaeer	Co-founder and former CEO of Vector Atomic; remained with IonQ after the October 2025 acquisition	Former DARPA program manager who initiated and managed more than ten quantum programs, including in positioning, navigation and timing; member of the National Quantum Initiative Advisory Committee [FACT]; Vector Atomic held more than $200 million in US government contracts, including work for the X-37B [CO-STATED]	GPS-independent timing; the quantum sensor in orbit; R-9
Thomas Sonderman	CEO, SkyWater Technology (IonQ subsidiary)	Leads a DMEA-accredited Category 1A Trusted Foundry serving federal defence programs; built SkyWater’s US government business [FACT]	Trusted fabrication; R-3
Bill Dunlap	Senior Vice President, Global Architecture	Former acting principal deputy Chief Information Officer at the Department of Defense; earlier DARPA’s CIO, running classified IT and special-access program systems; Air Force Security Forces veteran [FACT]	The Trusted Quantum Stack (R-3)
Leslie Kershaw	Chief Information Security Officer (from IonQ Capella)	More than twenty years in cyber operations, from the military to industry [CO-STATED]	Securing the stack and its orbital operations (R-3, R-9)

Table 7.4. Roles as listed by IonQ and SkyWater as of September 2026; backgrounds from IonQ releases, IonQ governance pages, quantum.gov and trade press (Appendix C). These individuals did not review, contribute to or endorse this report, and nothing here speaks for them or for IonQ.

The builders behind the bench. Two leaders from acquired companies are not national-security figures and are not presented as such. They are included because the national-security case rests on what they build [INFER].

Name	Role at IonQ	Background	Why they matter here
Dr. Chris Ballance	President, Quantum Computing	Co-founder of Oxford Ionics, whose electronic qubit control on chip IonQ acquired in 2025 [FACT]	The processor and control layers; Superion 256 and the Walking Cat roadmap the decoder serves
Dr. Mihir Bhaskar	SVP & General Manager of Quantum Technologies, SkyWater	Co-founder and CEO of Lightsynq; former Harvard quantum-networking researcher who launched and led the AWS Center for Quantum Networking; led the 2020 demonstration of memory-enhanced quantum networking [FACT]	Quantum interconnects and repeaters — the entanglement path in Section 9A and the modular scaling of IonQ’s machines

Table 7.4b. Neither has a national-security background; both are vital to the layers the national-security case depends on. Titles as listed by IonQ, September 2026. Neither reviewed or endorsed this report.

7.5 Where the case is bounded

• Compute benchmarks. Quantinuum, IBM and Google each lead on specific fidelity, error-correction or roadmap measures. IonQ has its own marker: 99.99% two-qubit gate fidelity, achieved in 2025 and described by the company as a world record [CO-STATED]. The case for IonQ is breadth of owned integration, not supremacy on every metric [INFER].
• Company self-description. “The world’s leading full-stack quantum platform and foundry” is IonQ’s own phrase [CO-STATED]. This paper does not adopt it; it adopts the narrower, checkable claim in KJ-5.
• Competitors can assemble. IBM could add owned fabrication when its funded foundry operates; Quantinuum draws on two accredited suppliers; a defence integrator could buy best-of-breed components [INFER].
• Simulation, not hardware. The decoder result assumes the error rates and 1–5 ms cycle times of an architecture that is still a blueprint; independent commentators note it is not a hardware demonstration [FACT]. It removes a bottleneck on paper; it does not yet show a fault-tolerant machine [INFER].
• Merchant-foundry tension. The FTC split 1–1 on a settlement that would have required IonQ not to disadvantage rivals using SkyWater [FACT]. A trusted national foundry owned by one competitor must remain credibly open to others [ARG].

7.6 How the IonQ case could fail

Failure mode	What it would look like	Consequence for this paper
Integration strain	Slippage in Superion 256 deliveries or SkyWater merchant service	Weakens the owned-stack advantage in practice
Architecture risk	Electronic qubit control does not scale as planned	Processor layer reverts to parity or worse
Concentration	One company becomes a single point of failure for a national stack	Argues for multi-award design, which Section 10 adopts
Customer conflict	Rivals avoid SkyWater	Trusted-foundry capacity underused nationally
Capital dependence	Acquisition pace outruns organic revenue	Execution risk rises with scope
Decoder gap	Real hardware noise or cycle times miss the simulation’s assumptions	Fault-tolerance timeline slips; the orbital argument in 9A.4 weakens

7.7 What IonQ must prove in the next twelve months

The case is only as strong as its next milestones. This paper argues for IonQ; it also sets out, in advance, the evidence that would show the argument holding or failing [ARG].

Milestone	What would count as proof	Window	If it slips
Superion 256 in customers’ hands	First customer delivery of the 256-qubit system made at SkyWater, with on-chip electronic qubit control	2027 (company plan) [CO-STATED]	Processor-layer lead narrows; Section 7.6 integration risk rises
Decoder on real hardware	The single-CPU decoder keeping pace with live syndrome data from an IonQ system, not simulation	2026–27 (authors’ test) [WATCH]	Fault-tolerance timeline and the orbital argument weaken
Trusted foundry open to all	SkyWater continuing merchant and trusted-flow service to outside customers, visible in filings	Each quarterly report [WATCH]	The FTC concern in Section 7.5 gains force
A stack-level federal buyer	Any federal award spanning more than compute — or a Quantum Genesis application	From 19 October 2026 [WATCH]	The American version stays commercially led while China’s is state-backed
Space QKD network	A first dated payload, launch or on-orbit demonstration for the network announced in May 2025	Not yet dated [UNDISC]	The orbital lead remains assembled, not operating
Scale	Delivering 2026 revenue guidance of $450–460 million	10-K, early 2027 [CO-STATED]	Capital-dependence risk in Section 7.6 rises
Integration as one enterprise	Evidence against the mechanisms in Table 7.3: a common roadmap, shared architecture, managed dependencies, clear priorities and decision rights, combined customer offers, named accountability	Ongoing [WATCH]	The portfolio stays broad but federated; the case rests on assembly alone
Value realization — products	A product only the combination could build, beyond the 256-qubit chip (e.g., an integrated compute-network-timing system)	2026–27 [WATCH]	Integration adds breadth, not capability
Value realization — customers and contracts	A contract won because IonQ could bid several layers at once — above all a stack-level federal or sovereign award	2026–28 [WATCH]	The buyer case for one integrated supplier is unproven
Value realization — revenue	Disclosure that cross-unit offers generate revenue beyond the sum of the units	From 2027 filings [WATCH]	Growth is acquisition-driven, not integration-driven

8. Integration Versus Plurality: Three Scenarios to 2030

In one line. The outcome turns less on who builds the first fault-tolerant machine than on whose trusted stack it lands in — and every scenario has indicators observable this year.

Scenario	Description	Indicators to watch	Where it leaves the US
A. Coordinated scale	China’s cluster converts pilot lines, carrier clouds and demand plans into a fielded stack before American plurality converges	Origin pilot-line acceptance; repeat defence-SOE-led rounds; first Chinese logical-qubit result; Scenario Handshake contracts	Behind on infrastructure even if ahead on science
B. Plurality wins the physics	An American entrant reaches the Quantum Genesis 100-logical-qubit milestone first	Q Competition awards; logical-qubit demonstrations; IBM and Quantinuum roadmaps	Ahead on compute, but exposed if the machine sits on accessed components and foreign clouds
C. Integrated plurality	The US keeps multiple routes to fault tolerance and buys and trusts at least one owned stack end to end	A stack-level federal purchase; a trusted-stack framework; domestic cryogenic capacity	The configuration this paper recommends [ARG]

America’s fragmentation is not only a weakness. Many companies and many modalities give the United States competition, redundancy and options if one technical path fails — advantages a centralized model gives up [INFER]. The opportunity is not to recreate China’s system, but to gain the advantages of integration without losing those of the American model: at least one fully integrated, trusted stack, and a competitive field around it [ARG]. That is Scenario C, and it is why R-4 is multi-award.

The Series does not assign probabilities to scenarios. The indicators are chosen so that a reader can tell within twelve to eighteen months which path is being taken [ARG].

9. The Allied Dimension

In one line. The one concentrated point of leverage over China’s stack sits in three allied capitals, and it is depreciating on a schedule set in Hefei.

The global supply of dilution refrigerators is concentrated in vendors based in Finland, the United Kingdom and the Netherlands, and helium-3 derives almost entirely from tritium decay in North American stockpiles [FACT]. BIS listed Chinese distributors of dilution refrigeration in March 2025, confirming both sides regard cryogenics as a chokepoint [FACT]. China is also writing the rules for applied quantum at home: on 23 April 2026 State Grid Anhui led the founding of the country’s first power-quantum standardization technical committee [FACT] — domestic standards that will travel with Chinese grid-quantum products as they are exported [INFER]. Chinese localisation and a possible Russian isotope channel mean that leverage should be treated as contingent, not structural [INFER].

EO 14413 directs the Secretaries of State and Commerce to harmonise investment restrictions, research security and export controls with allies, and to align Pax Silica with its priorities [FACT]. It does not create an instrument for the three chokepoint states specifically, nor burden-sharing for the commercial cost they would bear [FACT]; [ARG]. Japan’s drone-mounted quantum magnetometer program and the Republic of Korea’s GPS-free quantum navigation line are the partner capabilities closest to fielding [FACT]. IonQ’s allied footprint — key distribution in Romania, ID Quantique in Switzerland, Oxford Ionics in the United Kingdom — makes it an unusually natural vehicle for allied stack work [FACT]; [INFER].

Italy shows what an allied answer can look like. At IonQ’s Investor Day at the New York Stock Exchange on 8 September, Italy’s Undersecretary for Technological Innovation, Alessio Butti, met IonQ’s leadership to discuss Italy’s quantum ambitions, and IonQ is reported to be part of the Q-Alliance, Rome’s initiative linking companies, universities and research centres in quantum [FACT] (Q-Alliance membership as reported). Two weeks later, Butti proposed an international framework joining AI, quantum, cybersecurity, energy and digital infrastructure, with cooperation extending to secure, interoperable satellites; “Regulating, investing and cooperating with allies is our line,” he wrote [FACT]. Set against the CAS–Rosatom memorandum and Rosatom’s BRICS and SCO offer, that is the outline of a competing allied bloc — and an allied, trusted-relay satellite path of the kind R-9 proposes [INFER].

9A. The Orbital Race: Compute, Keys and Constellations

In one line. Whoever owns the orbital layer first — compute, keys, timing, links and launch — decides where a space-based quantum accelerator lands. Over six months China’s orbital schedule advanced while the West’s quantum-in-orbit schedule slipped. This is an infrastructure race, not a physics race.

THE INFRASTRUCTURE RACE. No power put a quantum processor into orbit between 22 March and 22 September 2026 [FACT]. China spent those six months networking computers in orbit, loading them with twenty AI models, launching new compute satellites and announcing gigawatt-scale orbital computing for 2030 — while Boeing’s Q4S and Europe’s Eagle-1 quantum satellites slipped to 2027 [FACT]. The first space-based quantum accelerator will not create an advantage. It will inherit one — from whoever already owns the orbit it runs in [ARG].

Method. Two search passes in each of Chinese, English, Russian, Korean and Japanese, covering 22 March to 22 September 2026. The dated record, China’s quantum-satellite history, other nations’ status and the ecosystem comparison are in Appendix B.

 

Figure 4. China’s Qianfan broadband constellation more than doubled in 2026, with private rockets and sea launches now carrying batches. Starlink remains far larger — about 11,128 active satellites — but Qianfan, Guowang and China’s computing constellations are state-directed from the start (Appendix B).

9A.1 Three things the record shows

1. China is building “space data, computed in space” — 天数天算. The Three-Body constellation achieved inter-satellite networking in February and carried twenty AI models by July; Chaozhisuan-1 now identifies targets in orbit and downlinks only results; Shanghai’s Pearl plan targets gigawatt-scale orbital compute by 2030 [FACT]. That is the rack, link and launch layer into which quantum capability will later be installed [INFER].

2. Whoever owns the quantum satellite holds the keys. China is the only country known to have run intercontinental quantum key exchange from its own satellites with foreign partners — Austria, Russia and, via the 23-kg-payload Jinan-1, South Africa over 12,900 km [FACT]; [INFER] for “only”. Jinan-1 worked as a trusted relay: the satellite handles the keys it passes, so its owner sits inside the circle of trust [FACT]; [INFER]. Russia has no quantum satellite of its own; its 3,800-km link with China ran through Micius, which re-entered the atmosphere in January 2026 [FACT]. When Beijing offers quantum-satellite service through BRICS and the SCO, it is offering to sit inside its partners’ keys — and inside the Russia–China bloc, Beijing already owns the orbital quantum layer [ARG].

Micius is gone — expect Beijing to move fast on a replacement. With Micius retired in January 2026, China’s operating quantum-key capacity in orbit rests on Jinan-1, a single microsatellite launched in 2022 [FACT]; [INFER]. Five pressures point to an aggressive launch schedule [INFER]: (1) a stated deadline — the higher-orbit quantum satellite was targeted for launch readiness by the end of 2026 [FACT]; (2) a five-year-plan target — the integrated space-ground quantum network is written into the plan, and a gap in orbit is a gap in that plan [FACT]; (3) promises to partners — the Russia link ran through Micius, and quantum cooperation offered through CAS–Rosatom, BRICS and the SCO needs a satellite to deliver it [FACT]; [INFER]; (4) a window to be first again — the American and European quantum-networking satellites have slipped to 2027, so a Chinese launch in 2026 or early 2027 would keep China the only country offering satellite quantum keys to foreign partners [FACT]; [INFER]; and (5) the means — a launch cadence that now includes reusable boosters, sea launches and small rockets claiming thirty flights a year [FACT]. Its absence is a gap today; its replacement is the next orbital indicator to watch [ARG].

3. America leads in scale; China leads in integration. Starlink is roughly two-thirds of everything in orbit, and the SDA’s laser-linked military transport layer is deploying [FACT]. But America’s largest constellation is a commercial service the military buys from, and its quantum-satellite effort is a single Boeing-funded experiment now due in 2027 [FACT]. Washington’s scepticism of QKD for national security systems is defensible [FACT]; declining China’s relay race is not a strategy. The way past the key-holder problem is entanglement — links in which no satellite holds the key — and that is the generation the United States should lead [ARG].

9A.2 The orbital stack, compared

Appendix B.5 sets China, the United States and IonQ side by side across seven orbital layers. In short: China leads in compute and quantum keys in orbit; the United States leads in constellations and launch; IonQ already flies a quantum sensor, laser terminals and radar, and has announced — but not yet flown — a quantum-key network [FACT]; [INFER].

9A.3 What it means in the theater

The kill chain moves into orbit — target identification on board cuts sensing-to-targeting from hours to minutes and reduces dependence on jammable ground stations [FACT]; [INFER]. Every orbital computer is a harvest-now target — only as sovereign as the links that command it and the clock that times it [ARG]. Space data is already a theater weapon — the Wall Street Journal reported Chinese imagery enabled an Iranian strike on US forces in July; China denies it; the report itself is [FACT]. The rules are about to be contested — Washington’s disclosure of weapons in orbit gives Beijing and Moscow momentum at the UN in November, and a placement treaty could freeze an orbital layer China has built and America has not [FACT]; [ARG].

9A.4 Where IonQ fits

IonQ is the only American quantum company that has both announced a space quantum network and already flies hardware in orbit. In May 2025 IonQ announced plans for a global space-to-space and space-to-ground quantum key distribution network, with the stated ambition of being the first company with both a quantum network and a quantum computer in space; on closing Capella in July 2025 it said it would begin developing that network [FACT]; [CO-STATED]; [INFER] for “only”. In orbit today: the Vector Atomic quantum inertial sensor on the X-37B, 84 Skyloom optical terminals flying on the SDA’s Tranche 1 Transport Layer, and Capella’s radar satellites under NRO contract [FACT]. On the ground: ID Quantique, the Chattanooga entanglement network and Lightsynq’s quantum-memory technology for repeaters [FACT]. And on paper: trapped ions that need no millikelvin refrigeration, and a decoder that runs on one CPU — the two heaviest burdens of a fault-tolerant machine in orbit, addressed [FACT]; [INFER]. IonQ’s architecture vice president puts energy-to-solution alongside time and cost as the company’s ‘North Star’ [CO-STATED] — the right measure for anything that must run on a satellite’s power budget [INFER]. That is the American orbital quantum stack: announced, assembled and partly flying — a quantum sensor and classical laser links and radar in orbit today; no quantum-communications or quantum-computing payload yet [ARG].

The line kept. IonQ has announced the space QKD network and the space-computing ambition; it has not flown quantum-communications or quantum-computing hardware for them, and no launch date was found [FACT]; [UNDISC]. Confirmed: IonQ hardware is on the US military’s laser-linked transport layer. Skyloom’s optical terminals flew on York Space Systems satellites for the Space Development Agency’s Proliferated Warfighter Space Architecture, launched 16 July 2026 — the second SDA deployment carrying the terminals — and Skyloom delivered 42 terminals for York’s first orbital plane of the Tranche 1 Transport Layer in March 2025, with a second batch of 42 for Plane 2 in progress [FACT].

10. What the Administration Should Consider Immediately

In one line. Nine actions, each checked against every instrument in Section 5.2, and each outside them — one tied to the summit and the November truce expiry, and eight that close exposures or buy, trust, secure, measure and extend the stack, on the ground and in orbit.

Method. Each action below was tested against EO 14413, EO 14412, EO 14409, the CHIPS quantum awards, the Quantum Genesis Q Competition, the National Quantum Initiative and current export controls. An action is included only if none of them already directs it [ARG]. Every procurement action is framed as competitive, milestone-based and open to any qualified American firm; where this paper names IonQ, it names it as the reference candidate to be tested, not as a sole source [ARG].

R-1. Carve quantum into the truce, before 10 November

The action. In any extension of the Busan truce agreed at or after the 24 September summit: (a) make no concession on quantum export controls; (b) if the Affiliates Rule suspension is extended, exclude from the extension any affiliate at least 50% owned by an Entity-Listed quantum parent; (c) keep quantum outside any Board of Investment liberalisation.

Why now. BIS stayed the Affiliates Rule for one year under the truce; the stay ends on 9 November 2026 and the rule returns on 10 November absent an extension [FACT]. The rule would automatically extend Entity List restrictions to subsidiaries at least 50% owned by listed parents [FACT]. The Chinese quantum cluster is built from exactly such parents — CAS institutes, USTC, QuantumCTek, Origin Quantum — and their spin-outs and investment vehicles [FACT]; [INFER]. In the week of the summit, Beijing has been pressing Washington to extend further its suspension of the rule barring thousands of Chinese firms from US advanced technology; the truce extension had not been agreed, and US officials described an acceptable extension of roughly three to six months, phased to verify compliance [FACT]. This is Beijing’s own technology ask — which is why quantum should be carved out of any answer to it [ARG].

Why it is outside current programs. EO 14413 addresses allied alignment, not the bilateral truce; the suspension covers all sectors without distinction [FACT].

Owner and horizon. The President; Commerce (BIS); USTR. Before 10 November 2026.

R-2. Close the quantum-cloud gap

The action. Direct Commerce to open a review under its information and communications technology supply-chain authority of US-person use of cloud services operated by Entity-Listed Chinese quantum firms, and, immediately, bar federal contractors and grantees from running federally funded work on them.

Why now. Origin Quantum is Entity-Listed, yet its cloud has served users in more than 160 countries with Americans reported among the most active [FACT]. Export controls govern what leaves the United States; they do not govern American circuits, data and benchmark workloads flowing into a listed firm’s machine [INFER]. Each such job improves the listed firm’s calibration data and user base [INFER].

Why it is outside current programs. No instrument in Section 5.2 addresses inbound use of foreign quantum clouds [FACT].

Owner and horizon. Commerce; OMB for federal contract and grant terms. Order within 30 days.

R-3. Create a Trusted Quantum Stack designation

The action. Direct the Department of Defense, with NSA, to define a trusted-stack standard covering fabrication and packaging, control electronics, firmware and operating system, cryogenic and photonic supply, and cloud operations — and to make it a qualifying condition for Defense and intelligence quantum purchases from FY2027.

Why now. Trusted accreditation today stops at the foundry door; DMEA accredits suppliers, not systems [FACT]. A processor fabricated in a trusted foundry can still run on imported control electronics, a foreign-made refrigerator and a commercial cloud [INFER]. China’s stack is controlled end to end by the state (Section 3); an American stack should be trusted end to end by design [ARG].

Why it is outside current programs. EO 14413 widens access to Defense-sponsored foundry resources; it creates no system-level trust framework [FACT].

Where IonQ qualifies. IonQ is, on the record in Section 7, the entity closest to meeting such a standard today, because it owns the accredited foundry and packaging and most layers above them [INFER]. The standard must be written so that teams — Quantinuum with its accredited suppliers, IBM with a future foundry, a defence integrator — can also qualify [ARG].

The openness condition. A trusted-stack designation that runs through a foundry owned by one competitor must make open access a written condition. IonQ has said SkyWater will continue to serve customers as a US-based merchant foundry, and the FTC’s 1–1 vote shows the question is live (Section 7.5) [FACT]. Written merchant-access terms should be part of the designation [ARG].

Owner and horizon. OSD (R&E); DMEA; NSA. Framework in 120 days.

R-4. Buy one integrated stack, not only a computer

The action. Award, under Other Transaction authority and existing research, development, test and evaluation funds, a multi-award prototype for one Sovereign Quantum Stack installation at a Defense or Energy site: compute, networking, resilient timing, post-quantum link protection with key distribution where the route qualifies, and documented trusted provenance — delivered as one accountable system.

Why now. The only milestone purchase in force is for computation, and it holds $2.5 million of $215 million [FACT]. China is building demand across the stack through the Scenario Handshake Plan [FACT]. The earlier an integrated American stack is under contract, the earlier a fault-tolerant processor from any American entrant has somewhere trusted to land [ARG].

Why it is outside current programs. QC-ADDS and Quantum Genesis buy computing capability; EO 14413’s advance-market-commitment language directs plans, not purchases, and none spans networking, timing and trusted provenance together [FACT].

Where IonQ qualifies. This paper proposes IonQ as the reference candidate: it is the one American company that could bid every element from owned assets [ARG]. The award should be multi-award at the first phase, so that a team assembled around another platform can compete on the same criteria — verified performance, error-correction progress, availability, supply-chain resilience, security, interoperability and lifecycle cost [ARG].

The automatic fallback. If the milestones in Section 7.7 slip — above all Superion 256 deliveries past 2027, or SkyWater merchant service disappearing from filings — the second phase of the award should move to multi-platform teams without a new decision [ARG].

Owner and horizon. Defense (OSD R&E, a service lead) or DOE Office of Science. Solicitation within 90 days.

R-5. Secure the cold: helium-3, isotopes and the Russian channel

The action. (a) Make a Defense Production Act Title III determination covering quantum-grade cryogenics and enriched isotopes; (b) direct the DOE Isotope Program to establish a priority allocation of helium-3 for domestic quantum and sensing production; (c) task Treasury and State to verify reported Rosatom isotope sales to Chinese buyers and, if confirmed, designate the parties.

Why now. Helium-3 has no substitute in dilution refrigeration and comes from tritium decay in North American stockpiles [FACT]. Chinese trade reporting says Rosatom’s isotope arm has agreed new supply of helium-3, germanium-72 and a silicon isotope to Chinese partners. That the claim was published is [FACT]; the transactions themselves are unverified [UNDISC]. If true, it routes around every allied licensing decision [INFER].

Why it is outside current programs. EO 14413 requires supply-chain plans and permits prize challenges; no Title III determination, allocation priority or sanctions tasking for this channel was identified in the public record reviewed for this report [UNDISC].

Owner and horizon. Defense (Industrial Base Policy); DOE; Treasury (OFAC); State. Verification tasking within 30 days; determinations within 90.

R-6. Measure China properly — and benchmark its machines

The action. (a) Direct the DNI to produce a National Intelligence Estimate on China’s quantum stack and conversion system within 90 days; (b) commission the PLA quantum procurement dataset that no one has built — the quantum equivalent of CSET’s 2,857-notice study of military AI procurement; (c) direct the national performance-assessment centre created by EO 14413 to benchmark cloud-accessible Chinese systems under controlled government access.

Why now. No Chinese quantum budget line is public, and funding-table comparisons mislead [UNDISC]. The procurement dataset is the only step that would move core judgments from structural inference to contract evidence [INFER]. Wukong-180 and Tianyan-287 are reachable from the open internet; the United States can test their claimed fidelities itself rather than cite them [FACT]; [ARG].

Why it is outside current programs. EO 14413 section 4(f) directs an assessment of the national security implications of commercial quantum computers generally, and its assessment centre is directed at quantum systems’ performance without a foreign-benchmarking mandate [FACT].

Owner and horizon. ODNI and the National Intelligence Council; DOE. Ninety days.

R-7. Answer Decree 841 with a talent lane

The action. Create an expedited adjudication lane for quantum specialists from allied countries, and for researchers leaving Chinese state programs, with counterintelligence vetting through the FBI’s quantum protection team that EO 14413 already expands.

Why now. China put into force on 15 September a decree permitting exit restrictions on people whose work touches export-controlled technology [FACT]. A country that restricts outbound movement of its specialists is signalling where it believes the binding constraint lies [INFER]. The American workforce measures in EO 14413 address federal hiring and domestic training [FACT].

Why it is outside current programs. No immigration-side quantum measure appears in any instrument in Section 5.2 [FACT]. This action involves security trade-offs on which reasonable officials differ; vetting is a precondition, not an afterthought [ARG].

Owner and horizon. State; Homeland Security; FBI. Policy in 90 days.

R-8. A chokepoint compact with Finland, the United Kingdom and the Netherlands

The action. Open a four-party understanding on quantum cryogenics: coordinated end-use controls; burden-sharing for the commercial cost borne by the three chokepoint states; and allied offtake commitments to expand capacity inside the Alliance. Invite Japan and the Republic of Korea as partners.

Why now. The leverage is real, concentrated and depreciating (Section 9) [INFER]. Unilateral national controls without burden-sharing produce substitution, not denial [INFER].

Why it is outside current programs. EO 14413 directs general harmonisation with allies; it names no instrument for the chokepoint states [FACT].

Owner and horizon. State; Commerce. Launch within 120 days.

R-9. Extend the stack to orbit

The action. Direct the Department of Defense, through the Space Force and the Space Development Agency, to (a) make quantum-safe link protection — post-quantum cryptography now, entanglement-based links where they qualify — and GPS-independent timing conditions of any federally procured or hosted orbital compute; (b) open a space-qualified quantum program line for orbital clocks, quantum-secure crosslinks and a quantum-processor-in-orbit pathfinder; (c) include the orbital layer in the Trusted Quantum Stack designation (R-3); and (d) adopt a trusted-relay rule — federal users, and allies through R-8, should not rely on quantum-key satellites owned by countries of concern — paired with an allied-owned entanglement-based quantum-relay pathfinder (Section 9A).

Why now. China has had computers in orbit since 2025, added new nodes on 20 September, and has a state plan for gigawatt-scale orbital compute by 2030 [FACT]. American orbital compute demonstrations are targeted for late 2027, and both the first American quantum-networking satellite and Europe’s quantum-key satellite have slipped to 2027 [FACT]. China has also run intercontinental quantum key exchange from its own satellites with Austria, Russia and South Africa, and is offering quantum cooperation through BRICS and the SCO [FACT]. The orbital layer is being built now; the quantum layer will be installed into it later (Section 9A) [ARG].

Why it is outside current programs. EO 14413 directs a civilian space plan from NASA and three Defense sensor projects, and DIU’s June initiative of up to $200 million transitions quantum sensors and clocks to the Joint Force [FACT]. None creates a program line for quantum in orbit, and none sets a link or timing condition for orbital compute [FACT].

Where IonQ qualifies. A quantum inertial sensor already flying on the X-37B, optical terminals on orbit, radar under NRO contract, and optical clocks under a DARPA award make IonQ a qualified competitor for (a) and (b) today [FACT]; [ARG].

Owner and horizon. Space Force; Space Development Agency; OSD (R&E). Direction in 90 days.

The nine actions at a glance

#	Action	Owner	Horizon	Scale anchor (existing comparable)	IonQ’s position
R-1	Quantum carve-in to the truce	President; BIS; USTR	Before 10 Nov	Regulatory: reinstating an existing BIS rule for one sector	Not applicable
R-2	Quantum-cloud gap	Commerce; OMB	30 days	Administrative: Commerce supply-chain review; contract and grant terms	Not applicable
R-3	Trusted Quantum Stack designation	OSD (R&E); DMEA; NSA	120 days	DMEA Trusted Foundry accreditation, extended from suppliers to systems	Closest to qualifying today
R-4	Buy one integrated stack	Defense or DOE	90 days	Quantum Genesis Q Competition, up to $215m; DIU sensing initiative, up to $200m	Proposed reference candidate; multi-award
R-5	Helium-3, isotopes, Russian channel	Defense; DOE; Treasury; State	30–90 days	CHIPS quantum letters of intent, $2.013bn; DPA Title III	Trapped ions less exposed to the dilution chokepoint
R-6	Net assessment and benchmarking	ODNI; DOE	90 days	CSET’s 2,857-notice military-AI procurement study	Not applicable
R-7	Talent lane	State; DHS; FBI	90 days	Administrative: adjudication priority plus FBI vetting	Not applicable
R-8	Chokepoint compact	State; Commerce	120 days	Diplomatic: four-party understanding	Allied footprint in UK, Switzerland, Romania
R-9	Extend the stack to orbit	Space Force; SDA; OSD (R&E)	90 days	DIU sensing initiative, up to $200m; existing SDA transport-layer buys	Quantum sensor on X-37B; terminals and radar on orbit; clocks under DARPA award

Scale anchors are existing comparable programs, not cost estimates [ARG].

Part IV’s five charter asks for the AI Force — infrastructure, migration clocks, evidence, weights in transit, and government as customer — remain the companion agenda and are not repeated here.

11. Watch List

Date or trigger	What to watch	Tier
25 September 2026	Post-summit addendum to this report, updating R-1	[WATCH]
23–25 September 2026	Xi state visit; any truce extension and its treatment of the Affiliates Rule; Quantum World Congress, College Park	[WATCH]
1 October 2026	Mobilization Law in force; implementing regulations; any designated-entity lists	[WATCH]
19 October 2026	Quantum Genesis Q Competition applications close; appropriations follow-through	[WATCH]
9–10 November 2026	Affiliates Rule stay ends 9 November; rule returns 10 November absent an extension	[WATCH]
December 2026	Updated National Quantum Strategy due under EO 14413	[WATCH]
Any time	First published application of Decree 841 to a quantum specialist	[WATCH]
Any time	Origin Quantum IPO filing and syndicate; next Wukong or Zuchongzhi generation; first Chinese logical-qubit claim	[WATCH]
2027	Superion 256 customer deliveries; IBM and GlobalFoundries foundries begin operation	[WATCH]
Mid-2027	Chinese local-debt resolution deadline — the fiscal test of the municipal model	[WATCH]
November 2026	UN disarmament meeting: space-weapons and placement rules	[WATCH]
End 2026 – early 2027	China’s higher-orbit quantum satellite — the replacement for Micius; launch readiness targeted for end-2026	[WATCH]
2027	SpaceX orbital-compute demonstrations; Boeing Q4S launch; Eagle-1 (February); Korean in-flight quantum-processor test	[WATCH]
Any time	Results from the X-37B quantum inertial sensor; Three-Body constellation’s next launches	[WATCH]
2026–27	IonQ’s real-time decoder run against live hardware syndrome data; any comparable Chinese decoding or logical-qubit result	[WATCH]
30 September 2028	EO 14413 deadline to field three Defense quantum sensor projects	[WATCH]
End 2026	Qianfan reaches its 324-satellite target; remaining SDA Tranche 1 transport launches	[WATCH]
Any time	China offering quantum-satellite key service to BRICS or SCO partners	[WATCH]
2030	Shanghai Pearl Constellation target for gigawatt-scale orbital compute	[WATCH]

12. Honest Concessions

1. Interest. The authors hold a long position in IonQ. Previous parts os this Series, by the same group, scored IonQ highest among full-stack platforms. Readers should weigh read material previously shared in the group's reports.
2. Ownership is not performance. Table 6 maps who owns what. It does not show that an owned layer is a better layer, and on several compute measures competitors lead [INFER].
3. Integration can be a liability. A single company spanning eight layers has inherent execution risk; the FTC’s split vote shows the merchant-foundry question is live [FACT].
4. China is more open than a fortress model suggests. Origin’s global cloud and the megaproject’s use of international reviewers slighly cut against a pure-denial reading [FACT].
5. China’s program may be smaller or more fragile than assessed. A program principal has suggested circulating figures might be inflated, and fiscal stress is documented, while others indicate the circulation figures are dramatically reduced [FACT].
6. Sources. Several China facts rest on trade press or single outlets, and this report was built mainly from English-language sources with Chinese primary text where cited [FACT].
7. Summit reporting moves fast. The truce terms and expiry date are drawn from reporting in the week before the summit and may change on 24 September [FACT].
8. The decoder is a simulation. The 22 September result decodes simulated workloads on IonQ’s own architecture; its assumptions have not been met in hardware, and a CPU-only decoder is, by design, easy for others — including Chinese teams — to reproduce once published [FACT]; [INFER].
9. Orbital claims are early and self-reported. China’s higher-orbit quantum satellite status is unconfirmed [UNDISC]. Chinese space-computing performance figures come from companies and state media; the economics of orbital compute are disputed by informed sceptics [FACT].
10. Policy recommendations are contestable. R-2 and R-7 in particular involve trade-offs between openness and security on which reasonable officials differ [ARG].
11. The de Masi parallel is an argument. Pairing a CEO’s acquisition strategy with a state program is the authors’ framing. Assembly is not integration: acquisition-led growth carries the execution and concentration risks in Section 7.6, and a company answers to its shareholders rather than to national priorities [ARG].

13. What Would Change This View

C1. An American competitor assembles an equivalently owned stack. IBM’s funded foundry operating alongside its compute and networking partnerships would remove IonQ’s uniqueness in Table 6, while leaving R-3 and R-4 intact.

C2. China publishes a verified logical-qubit or error-correction result. That would move the computing rung in Section 4 and strengthen Scenario A.

C3. China’s fiscal base forces retrenchment. Guidance-fund unwinds, corporate exits not absorbed by the state, or a milestone drought beyond eighteen months would weaken KJ-1 and KJ-3.

C4. The 2026 statutes stay dormant. No application of Decree 841 or the Mobilization Law to quantum personnel or firms within twelve to eighteen months would lower the weight placed on statute in Section 3.

C5. IonQ’s integration falters. Material slippage in Superion 256 deliveries or SkyWater merchant service, or no visible value realization from the combined businesses by 2028, would weaken KJ-5 and shift R-4 toward teams assembled around other platforms.

14. Conclusion

Part One — The Race Is for the Stack, on Earth and in Orbit

China is not racing to build one machine. It is racing to own everything a quantum machine depends on — the chips, the control, the refrigerators, the clouds, the customers, the networks, the keys and, now, the orbit [INFER]. It builds by decree: a national laboratory that runs the program, state carriers that own the networks, defence capital that funds the pilot lines, and statutes that hold its specialists and can mobilise its firms [FACT]. In the past six months it networked computers in orbit, launched a satellite that identifies targets on board, announced gigawatt-scale orbital computing for 2030, and remains the only country known to have run intercontinental quantum key exchange from its own satellites with foreign partners — sitting, by design, inside every key it relays [FACT]; [INFER] for “only”. What this report could not locate is a published Chinese logical-qubit result or a real-time decoder at scale [FACT]. Beijing’s strategy does not need one yet. It is building the house the breakthrough will move into [ARG].

A breakthrough does not decide this race. The stack it lands in does [ARG].

Part Two — Today, the American Stack Reached Its Hardest Layer

On 22 September 2026 IonQ showed that the hardest classical job in fault-tolerant quantum computing — correcting a machine’s errors in real time, across hundreds of logical qubits and millions of operations — can be done on a single ordinary processor, with the computation slowed by less than a third of one per cent [FACT]. It is a published simulation of IonQ’s own architecture, and live hardware is the next proof [UNDISC]. But it lands in a stack that already exists. The foundry is American and trusted: SkyWater, where IonQ’s first integrated 256-qubit chips were fabricated [FACT]; [CO-STATED]. The control is on the chip, from Oxford Ionics [FACT]. The keys and networks are ID Quantique’s and a US utility’s fibre [FACT]. The clocks and the quantum sensor are Vector Atomic’s — one of them flying on the X-37B [FACT]. The laser links are Skyloom’s, on the Space Development Agency’s military transport layer; the radar is Capella’s, under NRO contract; and IonQ has announced a space-to-space and space-to-ground quantum key network and the ambition to put a quantum computer in orbit [FACT]; [CO-STATED]. A trapped-ion machine that needs no millikelvin refrigeration and a decoder that needs one chip are, on paper, the two things a fault-tolerant accelerator in orbit could not do without [INFER].

That is Niccolo de Masi’s vision made physical. He had the audacity to believe an American company could match a state’s full-stack strategy, layer for layer — and in sixteen months of acquisitions and in-house research, IonQ assembled it, the American way: in public markets, under antitrust review, with trusted-supplier accreditation and shareholders to answer to — and staffed it with people who have run America’s space, intelligence, defence-research and trusted-manufacturing institutions [FACT]; [ARG]. For national security, that is the difference that matters. A stack whose silicon has provenance, whose keys are held by allies, whose timing does not depend on a jammable signal and whose links already ride the military’s own constellation is not a research portfolio. It is infrastructure [ARG].

Assembling the pieces was the first challenge. Making them operate as one system — and create value together that none could create alone — is the next, and it is the measure this paper will hold IonQ to (Sections 7.3 and 7.7) [ARG].

What remains is Washington’s part. Beijing’s stack has a state customer; America’s does not yet [FACT]; [ARG]. Keep quantum inside the Affiliates Rule before it snaps back on 10 November. Designate a Trusted Quantum Stack. Buy the first one. Carry it into orbit before China’s orbital layer becomes the default. And test IonQ, in the open, against every competitor that believes it can match it — against the milestones this report has set out in advance [ARG].

China built its stack by decree and is carrying it into orbit. Under Niccolo de Masi, an American company built the same stack by conviction and acquisition — and today showed it can carry the hardest layer of a fault-tolerant machine on a single chip. The United States does not need to copy Beijing. It needs to buy the stack it already has, trust it, and put it in orbit first.

Appendix A — Entity Register, Mirrored

China	Role	US counterpart (layer)
Hefei National Laboratory	Administers the megaproject competition; largest performer	No equivalent single administrator; DOE, NSF, DARPA fund separately
MOST	Issues megaproject guidelines	OSTP / National Quantum Coordination Office
Origin Quantum	Superconducting full stack; Entity-Listed	IBM, Google, Rigetti (processors); IonQ (owned stack breadth)
USTC / CAS centre	Zuchongzhi; Xiaohong chip; Entity-Listed	National laboratories; university centres
China Telecom Quantum Group	Tianyan cloud; controls QuantumCTek	Hyperscalers (access, not state-owned)
QuantumCTek	National quantum-secure networks; Entity-Listed since 2021	IonQ / ID Quantique; EPB (utility network)
Hyqubit	Trapped-ion computers and components	IonQ; Quantinuum
CETC	Dilution refrigerators (XS1000); research institutes Entity-Listed	Allied vendors in Finland, UK, Netherlands
Norinco	Led Origin’s pre-IPO round	Commerce (CHIPS equity stakes)
China Mobile	First HYQ-B100 customer; middleware investor	No state carrier equivalent
Beijing E-Town / Scenario Handshake Plan	Demand creation, 60+ firms	Quantum Genesis Q Competition (compute only)
CAS–Rosatom channel	Photonic and quantum cooperation, May 2026	Allied cooperation under EO 14413 section 9
Chaozhisuan; ADA Space / Zhejiang Lab; Shanghai Pearl plan	Compute in orbit; 2,800-satellite and gigawatt-scale plans	SpaceX (2027 demos); Axiom nodes; IonQ (links, radar, timing in orbit)

Appendix B — The Orbital Record

 

Figure 5. Dated orbital events, February–September 2026, from the sources in Appendix C.

B.1 The six-month record, 22 March–22 September 2026

Date (2026)	Actor	Event	Layer	Tier
30 Mar	UK (Bristol, Heriot-Watt)	SPOQC quantum-communications CubeSat launched	Quantum links	[FACT]
23 Apr	China (State Grid Anhui)	China’s first power-quantum standardization technical committee founded	Standards	[FACT]
6 May	Republic of Korea	Post-quantum cryptography pilot transition extended to defence, space, finance, telecoms and transport	Link protection	[FACT]
May	China (Origin Quantum)	Wukong-180, a 180-qubit processor, goes live on Origin’s global cloud	Compute (ground)	[FACT]
9 Jun	United States (SpaceX)	Investors told orbital-compute demonstrations are targeted for late 2027; up to one million data-centre satellites requested from regulators	Compute in orbit	[FACT]
17 Jun	Republic of Korea (INNOSPACE, Norma)	Agreement to fly a quantum processor on the HANBIT launcher and build a Space Quantum Computing Center; timing not set	Quantum compute in space	[FACT]
18 Jun	United States (Boeing)	Q4S quantum-networking payload qualified; launch moved to 2027	Quantum links	[FACT]
24 Jun	United States (DIU)	Up to $200m initiative to transition quantum sensors and clocks to the Joint Force	Sensing; timing	[FACT]
18 Jul	China (USTC, WAIC)	USTC president: computing’s physical limits answered by quantum and space computing; quantum-intelligence alliance and space-computing committee launched the same day	Strategy	[FACT]
Jul	China (Zhejiang Lab)	Three-Body constellation adds two satellites; 20 AI models now deployed in orbit	Compute in orbit	[FACT]
Jul	ESA	Eagle-1 quantum-key satellite now slated for February 2027	Quantum links	[FACT]
29 Jul	China (State Grid Anhui, Hefei)	People’s Daily reports Houdian substation’s 85 domestic quantum devices in grid service; diamond sensing material had been cut off by foreign suppliers and replaced domestically; power-flow calculations verified on Origin Wukong	Demand; sensing; communications; components	[FACT]
Aug	Japan	Michibiki-7 navigation satellite; defence request for drone-mounted quantum magnetometers	Timing; sensing	[FACT]
1–11 Sep	Russia (Rosatom)	Draft quantum programme to 2036 with export offer; cooperation offered to SCO; BRICS quantum forum made regular	Bloc formation	[FACT]
8 Sep	United States (DIU, Honeywell)	Quantum magnetic-navigation flight test over open ocean without GPS	Sensing; navigation	[FACT]
8–9 Sep	Italy (Butti)	Meets IonQ leadership at Investor Day, NYSE; US mission with quantum in focus	Allied cooperation	[FACT]
11 Sep	United States (SpaceX)	CFO: orbital compute could be competitive with ground data centres as early as next year	Compute in orbit	[FACT]
12–13 Sep	China (Shanghai; Beijing–Tianjin–Hebei)	Pearl Constellation: gigawatt-scale orbital compute by 2030; four-jurisdiction space-computing corridor	Compute at scale	[FACT]
14–16 Sep	United States	Weapons in orbit confirmed; Beijing urges Washington to stop preparing for war in space	Theater	[FACT]
20 Sep	China (Chaozhisuan; Pengcheng)	Chaozhisuan-1 identifies targets in orbit and downlinks results; Pengcheng satellite fuses 5G base station, AI compute, laser and microwave links	Compute and links	[FACT]
22 Sep	Italy (Butti)	Op-ed proposes an international framework for AI, quantum, cybersecurity, energy and secure interoperable satellites	Allied cooperation; standards	[FACT]

B.2 China’s quantum satellites, 2016–2026

Year	Satellite / result	What it proved	Tier
2016	Tiangong-2 cold-atom clock: first cold-atom clock operated in orbit, running for almost three years	Quantum timing works in orbit	[FACT]
2016–17	Micius: 1,000-kilometre-class entanglement distribution, key distribution and teleportation; Beijing–Vienna intercontinental key exchange over 7,600 kilometres	Satellite quantum links work at continental scale	[FACT]
2022	Jinan-1 launched — the first quantum microsatellite	Quantum payload cut to about 23 kg from about 250–300 kg on Micius; ground stations cut from about 13 tonnes to about 100 kg	[FACT]
2023–24	Russia–China quantum-protected channel of about 3,800 km, via Micius, with a Russian ground station near Moscow	The Russia–China bloc’s only long-range quantum link runs through a Chinese satellite	[FACT]
2025	Jinan-1 real-time key distribution to mobile stations in six Chinese cities and South Africa; up to 1.07 million secure bits in one pass (Nature, March 2025)	Satellite quantum keys move from experiment to deployable service	[FACT]
Jan 2026	Micius re-enters the atmosphere after nearly ten years in orbit	China’s operating quantum-key capacity in orbit falls to Jinan-1; the Russia–China Micius channel ends	[FACT]
2026 target	A medium-to-high-orbit quantum satellite at roughly 10,000 km altitude, payload performance about ten times higher, launch readiness targeted for end-2026	Global coverage from a few satellites; linking moving targets	[FACT] (2024 statement); [UNDISC] (current status)
Plan	A “quantum constellation” of higher-orbit satellites and practical low-orbit microsatellites connected to a national fibre network of more than 12,000 km	Integrated space-ground quantum network, a five-year-plan target	[FACT]

B.3 Everyone else (status at 22 September 2026)

Actor	Quantum satellite status (22 September 2026)	Tier
European Union / ESA	Eagle-1 key-distribution satellite slated for February 2027; EuroQCI and SAGA to follow; Spain’s geostationary QKD-GEO prototype in development	[FACT]
United Kingdom	SPOQC CubeSat launched 30 March 2026, in commissioning	[FACT]
Singapore–UK	SpeQtre entangled-photon CubeSat launched late 2025	[FACT]
United States	Boeing Q4S entanglement-swapping satellite, 2027; NSA and NIST regard QKD as impractical for national security systems and prioritise post-quantum cryptography	[FACT]
Japan	Satellite quantum cryptography funded through JAXA’s Space Strategy Fund; physical-layer key sharing demonstrated from the ISS in 2024; no dated national quantum satellite	[FACT]; [UNDISC]
Republic of Korea	Post-quantum migration extended to space and defence; no quantum satellite announced	[FACT]
Russia	No quantum-communications satellite of its own; a CubeSat link between Moscow and Kislovodsk targeted for 2030	[FACT]

B.4 The satellite ecosystem

Element	United States	China	Tier
Broadband mega-constellations	Starlink: about 11,128 active satellites of 16,561 active worldwide — roughly two-thirds of everything in orbit (September 2026)	Qianfan: 256 satellites after launches on 15–16 September; 324 targeted by end-2026; plan of 648, then 1,296, then more than 15,000. Guowang: 136 at end-2025	[FACT]
Military transport layer	SDA Proliferated Warfighter Space Architecture: laser-crosslinked Tranche 1 transport satellites; a York-built batch carrying IonQ’s Skyloom terminals launched 16 July 2026, further batches no earlier than Q3 2026	Space-ground integrated network as a five-year-plan target; state operators	[FACT]
Compute in orbit	Demonstration nodes (January 2026); SpaceX demonstrations late 2027; filing for up to one million data-centre satellites	Three-Body and Star-Compute constellations operating; 2,800-satellite and gigawatt-scale plans	[FACT]
Launch	World lead in cadence and reusability	First orbital booster recoveries on two separate pathways in July and August 2026; record sea launch of nine satellites on 16 September; Lijian-1 claiming 30 launches a year	[FACT]
Quantum payloads	X-37B quantum inertial sensor (flying); Q4S 2027	Jinan-1 (Micius retired January 2026); higher-orbit quantum satellite targeted	[FACT]
Stated bottlenecks	Launch slips on SDA batches	Qianfan’s operator cites launch capacity, tracking resources, and low-cost mass production of chips, payloads, antennas and terminals	[FACT]

B.5 The orbital stack, compared

Orbital layer	China	United States	IonQ
Compute in orbit	Operating since 2025; networked; 20 AI models; new nodes September 2026	Demonstration nodes launched January 2026; SpaceX demonstrations late 2027	None announced
Optical and laser links	Operating, combined with compute on one spacecraft	Operating at scale (commercial constellations)	Operating — Skyloom, 84 terminals on orbit
Quantum links	Micius retired January 2026; Jinan-1 microsatellite in orbit; higher-orbit satellite targeted for end-2026 readiness	SPOQC (UK) in orbit; Q4S and Eagle-1 in 2027	Announced — global space-to-space and space-to-ground QKD network (May 2025); no flight hardware yet
Quantum sensing in orbit	Cold-atom clock flown on Tiangong-2 (2016)	Strategic-grade quantum inertial sensor on the X-37B since August 2025	Flying — the X-37B sensor was built by Vector Atomic, now IonQ
Timing independent of GPS	BeiDou; quantum clocks a plan target	GPS; DIU clock transition initiative	Contracted — Vector Atomic optical clocks under DARPA award
Radar and imaging	Target identification in orbit claimed	Commercial and national systems	Operating — Capella, under NRO contract
Launch cadence	Lijian-1 alone claims 30 launches a year, booked into late 2027	World lead	Not applicable

Table B.5. Status as reported [FACT]; “operating”, “flying”, “contracted” and “announced” are the authors’ summary of cited sources [INFER].

Appendix C — Primary Sources

United States government. The White House, Executive Order 14413, Ushering in the Next Frontier of Quantum Innovation, 22 June 2026, whitehouse.gov/presidential-actions/2026/06/ushering-in-the-next-frontier-of-quantum-innovation · EO 14412 and EO 14409, Federal Register (91 FR 38483; 91 FR 34565) · NIST, CHIPS quantum letters of intent, 21 May 2026 · DOE Office of Science, Quantum Genesis Q Competition, 17 September 2026 · Federal Register, Additions of Entities to the Entity List, 14 May 2024 (2024-10485) · BIS final rule adding QuantumCTek and others, 26 November 2021 · BIS suspension of the Affiliates Rule, 10 November 2025 · DMEA accredited-supplier list, September 2026 · National Quantum Initiative supplements to the President’s Budget, quantum.gov.

Companies. IonQ, IonQ Completes Acquisition of SkyWater Technology, 31 July 2026, and Form 8-K exhibit (sec.gov) · IonQ, second-quarter 2026 results, 5 August 2026 · IonQ, full-year 2026 guidance, 8 September 2026 · IonQ Form 10-K FY2025 · Origin Quantum, OriginQ Wukong 180, May 2026 (originqc.com) · China Telecom Quantum Group, Tianyan: Cloud services with quantum advantage, arXiv 2512.10504 · QuantumCTek 2025 annual report, Shanghai Stock Exchange · Guo et al., A site-resolved two-dimensional quantum simulator with hundreds of trapped ions, Nature 630 (2024).

Trade and research press. Quantum Computing Report: Wukong-180 (14 May 2026); Origin Pilot (8 March 2026); IonQ–SkyWater completion (31 July 2026) · The Quantum Insider: Origin Pilot (27 February 2026); industry response to EO 14413 (23 June 2026); Chinese quantum companies (May 2026); Scenario Handshake Plan (16 September 2026) · Bloomberg, FTC split on IonQ–SkyWater, 29 July 2026 · MarketScreener, FTC statements, August 2026 · Entangled Future, Hyqubit profile · 36Kr, Hyqubit Series A · PostQuantum, Origin Quantum profile and Origin Pilot analysis; China’s quantum supply chain · Quantum Zeitgeist, China quantum companies guide, July 2026 · The Qubit Report, Wukong-180, 9 May 2026.

Summit and truce. Asia Society Policy Institute, What to Watch at the Xi–Trump Summit, September 2026 · Brownstein, Expectations for the Trump–Xi September Meeting, September 2026 · IBTimes, summit and truce, September 2026 · Letter of Senator Wyden and colleagues on the Affiliates Rule suspension, 12 November 2025.

Analysis. ASPI Critical Technology Tracker · RUSI on Chinese quantum supply-chain localisation, 2026 · Rhodium Group, China’s Financial and Fiscal Decay, March 2026 · CSET, Pulling Back the Curtain on China’s Military-Civil Fusion · Chinese Academy of Sciences, announcement of the CAS–Rosatom memorandum, 20–21 May 2026 (cas.cn).

Orbital layer, 22 March–22 September 2026. Science and Technology Daily and Xinhua on Three-Body constellation networking, 12–13 February 2026; Xinhua and CLS on its July 2026 expansion · China Daily on Qwen3 in orbit, 28 January 2026 · The Paper on the July 2026 World AI Conference · DIU, From Lab to Orbit Within Months, July 2025; Quantum Computing Report and Spaceflight Now on the X-37B OTV-8 launch, August 2025; Boeing on OTV-8 experiments, 2026 · DIU quantum sensing and timing initiative, 24 June 2026; DefenseScoop on the MagNav flight test, 8 September 2026 · ESA Connectivity and Secure Communications, EAGLE-1, July 2026 · University of Bristol on SPOQC, March 2026 · Newsis on Korea’s post-quantum pilot transition, 6 May 2026 · iXBT on Russia’s 2030 satellite quantum link; CNews and Izvestia on the Russia–China Micius link · JAXA Space Strategy Fund, satellite quantum cryptography theme; Nikkei xTECH on NICT satellite plans · Reuters on SpaceX orbital compute, 9 June 2026. Orbital layer, 22 August–22 September 2026. 21st Century Business Herald, Sohu, FX168 and Tencent News on the Lijian-1 Y18 launch and Chaozhisuan-1, 20–21 September 2026 · 36Kr and Tencent News on the Pearl Constellation plan and the Beijing–Tianjin–Hebei corridor, September 2026 · Xinhua on China’s first space-computing constellation, 15 May 2025 · 21st Century Business Herald on the July 2026 World AI Conference · Spacemapper daily, 8 September 2026 · CAS, China Science Daily, 19 August 2024 (Pan Jianwei on the higher-orbit quantum satellite) · CNN, CBS, NBC/Reuters and AFP on US weapons in orbit, 15–16 September 2026 · Global Economic (Korea) and Pulse 2.0 on SpaceX orbital compute, September 2026; Reuters, 9 June 2026 · Boeing and SpaceNews on Q4S, June 2026 · Atomic Energy 2.0 and TASS on Rosatom’s quantum programme, SCO offer and BRICS forum, September 2026 · SatNews, The Quantum Insider and Asia Economy on INNOSPACE–Norma, June 2026 · Cabinet Office of Japan, space policy notices, August–September 2026 · Axiom Space on orbital data-center nodes, January 2026.

Quantum satellites and ecosystem. Li et al., Microsatellite-based real-time quantum key distribution, Nature 640 (2025), and CAS, Stellenbosch University and EurekAlert releases, March 2025 · PostQuantum analyses of Jinan-1 and of China’s quantum networking programme, 2026 · Shanghai municipal government on the higher-orbit quantum satellite, September 2024 · ESA and ESA CSC on EAGLE-1, 2026 · Deloitte Japan on QKD-GEO · The Quantum Insider on SpeQtre, December 2025 · Tencent News, Sina Finance, Economic Observer and IT Home on Qianfan launches, May–September 2026 · Next Spaceflight on SDA Tranche 1 transport launches, 2026 · StockTitan summary of IonQ’s Skyloom terminal count after the 16 July 2026 launch · PLA Dual-Use Technology & Mobilization Readiness Assessment (Starlink and booster-recovery figures).

Skyloom–SDA. Skyloom, Completes Deliveries of SDA’s Transport Layer Tranche 1 Optical Communication Terminals for York Space Systems’ Orbital Plane 1, Business Wire, 27 March 2025 · Via Satellite, 27 March 2025 · The Quantum Insider, IonQ’s Skyloom Optical Communications Terminals Reach 84 On-Orbit Installations, 24 August 2026.

Italy. Alessio Butti, op-ed in Milano Finanza, September 2026, as reported by Decode39, 22 September 2026 · Decode39, Italian tech undersecretary kicks off US mission in NY, with quantum in focus, 9 September 2026.

Summit and Chinese-language coverage, 21–23 September 2026. Xinhua and People’s Daily on the 23–25 September state visit · Xinhua curtain-raiser, 22 September · Global Views Monthly (Taipei) and Liberty Times on truce talks and the summit agenda, citing Reuters · Yahoo News Taiwan on Beijing’s request on the entity rule · Tencent News, 开局之年看中国 (Micius re-entry; Wukong usage), September 2026 · Satellite encyclopedia (huijiwiki) on the 26 January 2026 re-entry · Chinese Academy of Sciences / Xinhua on Jiuzhang-4, 13–14 May 2026.

Houdian substation. People’s Daily, 探访国内首座量子应用示范变电站 (visit to China’s first quantum application demonstration substation), 29 July 2026, via Eastmoney · Xinhua, 我国首座量子应用示范变电站建成投用, 29 November 2024 · People’s Daily, 29 November 2024 (print edition) · China National Radio, Anhui, December 2024 (State Grid statement on the “power + quantum” innovation consortium) · Anhui News, 29 November 2024 (State Grid Anhui market estimate) · TipRanks, 18 August 2026 (English-language relay).

National-security bench. IonQ, IonQ Appoints General John W. “Jay” Raymond to Board of Directors, 29 September 2025 · IonQ, IonQ Announces IonQ Federal, September 2025, and governance pages (Cardillo) · IonQ, IonQ Expands Engineering Leadership Team, Hiring Rick Muller, 21 July 2025 · Investing.com on Dean Acosta’s appointment, September 2025 · quantum.gov, National Quantum Initiative Advisory Committee (Abo-Shaeer), and HPCwire / The Quantum Insider on the Vector Atomic closing, October 2025 · SkyWater and IonQ releases (Sonderman), 2026 · GovConWire on Bill Dunlap, April 2026 · IonQ release on CIO and CISO appointments, 14 January 2026 (Kershaw) · IonQ, Lightsynq acquisition releases, May–June 2025, and Harvard Innovation Labs profile (Bhaskar) · IonQ company page and Q3 2025 earnings call (Ballance) · IonQ Q3 2025 earnings call transcript.

IonQ primary sources. IonQ, IonQ Announces Plans for First Space-Based Quantum Key Distribution Network, 7 May 2025, and Capella completion 8-K exhibit, 15 July 2025 · IonQ Vector Atomic completion 8-K exhibit, 7 October 2025, and Form ARS FY2025 (acquisition date and consideration) · IonQ Skyloom completion release, 28 January 2026 · PrivSource deal summaries (Skyloom SDA-qualified terminals; ID Quantique; Lightsynq) · Liu et al., In-orbit operation of an atomic clock based on laser-cooled 87Rb atoms, Nature Communications 9 (2018).

IonQ decoder. IonQ, IonQ Demonstrates Industry’s First End-to-End Real-Time Quantum Error Decoder, 22 September 2026 (quotes from Nicolas Delfosse and John Gamble; company boilerplate on 99.99% two-qubit fidelity) · Ye, Maksymov and Delfosse, Real-time decoder for a MegaQuOp quantum computer using a single CPU, arXiv 2608.25027 (v1 25 August, v2 3 September 2026) · Quantum Computing Report on the result · Free Quantum Computing commentary noting simulation status · The Quantum Insider on the Walking Cat blueprint, 23 April 2026.

This Series and companion assessments. Friends of IonQ, An IonQ Solution for the AI Force, QTIS Part IV, 20 September 2026 · The Conversion Machine v3.0 · Quantum and the Alliance v1.4 · PLA Dual-Use Technology & Mobilization Readiness Assessment — Quantum Lane v3.0 with Final Revision Addendum, 16 September 2026.

Disclosure

This report is written by Friends of IonQ, a group of researchers and analysts focused on IonQ activity, the quantum ecosystem at-large, and how quantum technology can enhance the human condition while fortifying the national security of the United States and its allies. Its members hold a long position in IonQ. Neither Friends of IonQ nor any individual member receives any compensation or future consideration from IonQ. Where the group sees challenges the company needs to address, or issues that concern it, it raises them openly in its reports. This report is not investment advice, and it is not a solicitation to any government.