EuroHPC's six August quantum actions continue partnerships selected in 2022. The EU's platform-assessment roadmap is still a draft and is open for public comment.
On August 14, the EuroHPC Joint Undertaking, which buys Europe's shared supercomputers, announced "Six New Quantum Calls" worth €119 million: three €20 million platform-development actions for trapped-ion, superconducting and neutral-atom computing, each aimed at 1,000-qubit-class systems, plus quantum key distribution (€24 million), testing infrastructure (€20 million) and pilot fabrication (€15 million). The topics opened on the EU Funding and Tenders portal on August 13 and close on November 17.
The press release does not say where the six come from. In the Horizon Europe Work Program for 2026-27, each sits under a heading like "Call for the 2nd SGA for the Millenion FPA (trapped-ions)": second Specific Grant Agreements continuing six Framework Partnership Agreements, Millenion, OpenSuperQPlus, PASQuanS2, QSNP, Qu-Test and Qu-Pilot, selected competitively in the EU Quantum Flagship's 2021-22 round (proposals due January 27, 2022). The modality architecture behind the August money is four years old.
"Call" is not an invention: both parent work programs use the word themselves, and the portal topics are new listings with a real deadline. The announcement omits the lineage: no FPA, no second phase, none of the six program names beyond a Qu-Pilot testbed line. Read alone, it describes a fresh competition; the parent documents describe a continuation.
The platform choice dates to the Quantum Flagship's framework-partnership round of 2021-22, which produced Millenion (trapped ions) and OpenSuperQPlus (superconducting) from one topic on large-scale quantum computers, PASQuanS2 (neutral atoms) from quantum simulation, QSNP from quantum communications, and Qu-Test and Qu-Pilot from testing and pilot production; first-phase grants began in spring 2023. OpenSuperQPlus describes "two distinct project phases of 3.5 years each"; the 2026 work program instructs each August action to "implement the second 3.5 years of the action plan defined in the above FPA."
On paper, the instrument is closed. Each action is typed "Specific grant agreement awarded without call for proposals in relation to a Framework Partnership Agreement," and "the selected consortium will be invited to submit a proposal." EuroHPC's own work program, amended in April, describes the batch as a transfer: "the Commission will transfer the management of the six SGA2 linked to six existing FPAs," the signing process still to be agreed by its Governing Board.
The portal tells it differently. Launched August 13, all six topics are typed as standard Research and Innovation Actions: open submission, one expected grant each, and, as retrieved August 25, no clause restricting who may apply. The "selected consortium" sentence is gone; four of six launched texts still name their FPA, the fabrication action most plainly: "The action is to implement the pilot production elements of the Qu-Pilot Framework Partnership Agreement."
None of the launched documents states whether an outside consortium could submit to these topics and win.
Behind this timeline lies a straightforward explanation: the initial phase of grant funding is coming to an end. Millenion ends August 31, eighteen days after the topics opened; Qu-Test and Qu-Pilot end September 30. A November deadline, followed by evaluation and signature, reflects standard operational continuity across consecutive 3.5-year phases. These are continuation grants racing their predecessors' end dates.
One platform lane was created by a 2026 decision. PQC-06, opened June 2, is a conventional open research action for photonic quantum computing with no framework agreement behind it: a 100-qubit photonic NISQ processor by 2028, a full-stack machine by 2030 with an "indicative target" of 1,000 photonic qubits. The €10 million budget is indicative too; the work program notes it "does not preclude submission and selection of a proposal requesting different amounts."
Eligibility draws the same boundary. At work-program level, the photonic action admits entities from the UK, Canada, South Korea, New Zealand, Switzerland and Israel alongside member states, Iceland and Norway. The six phase-two actions limit participation to member states, Iceland, Norway and Israel, conditions the footnotes trace to the 2021-22 framework round. (Portal pages are inconsistent here; the work program is quoted.) The framework set was fixed in 2022. Photonics is the lane opened by present-day choice.
Semiconductor spin has neither a framework agreement nor a 2026 platform topic; its EuroHPC money moves through procurement and manufacturing. The EuroSSQ tender, for at least 16 physical spin qubits at SURF, closed October 24, 2025; as of late August 2026, EuroHPC has not announced a vendor. MeluXina-Q in Luxembourg, "at least 10 semiconductor spin qubits," closes September 1. Procurement built EuroHPC's installed base: six quantum computers procured and two co-funded through HPCQS per its portfolio page, machines at supercomputing centers in a wider on-premises turn. On the manufacturing side sit SPINS, the spin-qubit pilot line the Chips Joint Undertaking launched in April, and Quobly, which closed a €115 million Series A in June to industrialize spin qubits on STMicroelectronics' 300-millimeter line.
When the consortia were selected in early 2022, spin's demonstrated machine scale was small; it still is. The EU's draft platform assessment records multi-qubit operations in "arrays of up to 17 qubits" and notes the most recent demonstration involved quantum error correction, yet it gives spin one of five platform seats; spin's manufacturing evidence keeps outrunning its demonstrated scale. In the August texts reviewed here, diamond NV centers surface once, as a sensor example.
Three distinct factors explain this alignment, though the supporting evidence varies. Path dependence provides the strongest explanation, as framework agreements are designed for renewal, and these six projects align directly with the parameters established in their 2022 instruments. Technology maturity remains a plausible consideration, yet it lacks formal documentation; the available records offer no insight into how evaluators originally selected these consortia in 2021–22, and continuation grants proceed regardless of whether a new evaluation would yield the same outcome. Sovereignty presents a clear text-based rationale, albeit an indirect one, with eligibility restrictions and supply-chain requirements pointing to an industrial-policy framework that predates present qubit rankings. The sorting is real at the instrument level. What it says about the hardware is a reading, and the instrument histories are the confounder.
If the grants go to the incumbents, the money lands in known hands. Trapped-ion money would run through Millenion, coordinated by the University of Innsbruck, with AQT, Eviden, Toptica Photonics and QuiX Quantum among its members; eleQtron, the German trapped-ion startup, is absent from its public lists, funded instead through the national DLR and EPIQ channels. Superconducting money would fund Jülich-coordinated OpenSuperQPlus, already home to IQM, Alice & Bob, VTT, Bluefors and QuantWare. Neutral-atom money would fund Max Planck-coordinated PASQuanS2, with Pasqal, AQT and ParityQC among its partners; planqc, which leads Germany's national push for a 1,000-plus-qubit neutral-atom system at LRZ, does not appear in its public records. That bet runs on federal money outside the EU channel; Pasqal sits inside.
An open call asks who might win; a phase-two agreement asks who is already inside. Alice & Bob sits in both OpenSuperQPlus and France's PROQCIMA champion set.
The July 2025 Quantum Europe Strategy promised an EU Quantum Computing and Simulation Roadmap as the "evidence-based approach" that "will guide Europe's strategic decisions and help prioritize future public investments." The Quantum Flagship published first drafts of six roadmaps in late July and opened a public consultation; the deadline, extended to August 25, fell due the day this article was written, and the consultation remains open. The six phase-two topics launched August 13, mid-consultation. Those are the dates; the documented explanation remains the mundane one, expiring first-phase grants.
The draft roadmap, quoted as a draft whose final text may change, assesses five platforms: ions, superconductors, atoms, photons and spin. On convergence it is agnostic: quantum computing "may for many years remain a diverse sector rather than a 'winner takes all.'"
Sovereignty is its own thread. The 2021 framework topic admitted UK entities conditionally; no such slot appears in the 2026 phase-two texts, while the photonic action admits the UK outright. The draft roadmap's trapped-ion section warns that "US trapped-ion incumbents today rely on the EU supply chain for critical components" and that their "recent acquisitions signal strategic intent to extend US ownership" over the EU industrial base the Chips Joint Undertaking and the Quantum Europe Strategy "are mandated to protect." The passage names no acquirer. IonQ announced its $1.075 billion acquisition of the UK's Oxford Ionics in June 2025.
DARPA's Quantum Benchmarking Initiative declines to sort at all: it "is not a competition between performers." Entry is staggered; more teams are "likely but not guaranteed" to enter Stage B, it says. The 11 Stage B companies announced November 6, 2025 include silicon-spin teams Diraq, Quantum Motion, Silicon Quantum Computing and Photonic Inc., photonic Xanadu, and superconducting IBM and Nord Quantique; Rigetti, HPE and Alice & Bob were not among them. Mixed-modality qubit programs keep opening beside it. The contrast with Europe is temporal: architecture-open evaluation on a rolling clock, against a platform set fixed in 2021-22 and now funding its second half. Neither is an elimination tournament.
France's PROQCIMA is champion procurement: €500 million from defense agency DGA behind five companies picked in March 2024 (Alice & Bob, Pasqal, Quandela, Quobly, C12), with planned down-selection to three, then two, over roughly a decade. Germany's DLR Quantum Computing Initiative contracts builds across ion traps, NV-center diamond, neutral atoms, photonics and solid-state spin, with no superconducting build on its project list as of late August; its modality mix differs materially from the EU's.
The consultation closes August 25; MeluXina-Q's tender closes September 1; the photonic action closes September 30; the six phase-two topics close November 17. A EuroSSQ vendor announcement is pending, ten months after that tender closed. The final roadmap is due in 2026 under the strategy's action list, worth reading for two things: whether the published assessment ratifies or complicates the architecture inherited from 2022, and what the next tranche of platform money does with it.
The SCN Staff is a small AI editorial squad working under human direction. Each agent owns one job.
Scout does the research. It runs down primary sources and checks what's already been published, on SCN and everywhere else, before a story gets written. If a claim can't be traced back to a real document, Scout flags it.
Forge writes. It takes what Scout found and turns it into a draft, argument and sentences and all. Every SCN piece starts here, then gets sharpened.
Cipher handles search: the titles, descriptions, and keyphrase work that decides whether a good article ever gets found. Least glamorous job on the squad. Also one that matters more than it looks.
Pixel makes the visuals. Images, charts, the occasional diagram, all built to SCN's brand instead of pulled from a stock library. When something's easier to see than to read, it goes to Pixel.
Editorial judgment and the final call stay with the humans. So does the fact-checking.
