Intelligence Briefing deeptech

Deep Tech Briefing
July 31, 2026

The Dutch state pushes Chinese executives out of a Nijmegen chip maker, while Spain's Minoryx wins a rare-disease approval Europe refused it two years ago.

Europe spent this week deciding who gets to be trusted, and with what.

The Dutch government pushed the last Chinese executives out of a chip maker it does not own. A Spanish biotech won a recommendation for a drug the same committee turned down two years ago, and a quantum company published the benchmark that could falsify its own advantage claim.

One question sits under all three: what counts as verified, and who decides.

Lead

The Dutch State Empties Ampleon's Boardroom of Chinese Nationals

Ampleon, the Nijmegen RF chip specialist owned by China's Wuxi Xichanweixin Semiconductor, has removed all Chinese nationals from its executive and supervisory boards. Executive Jiaqi Wang departed, leaving CEO Dave Hartskeerl as the sole statutory director.

The company had already amended its articles to require a two-thirds supervisory board majority before any R&D leaves the Netherlands. Both changes followed talks with the Dutch government about technology transfer.

Nothing was nationalised and no minister issued a public order. Less than a year after The Hague intervened at Nexperia, foreign owners of Dutch chip assets are rewriting their own governance before being told to, which is a quieter instrument than a seizure and a more durable one.

(Bits&Chips)

Biotech & Life Sciences

Minoryx Wins the Approval Europe Refused It

The EMA's human medicines committee adopted a positive opinion on July 23 for Nezglyal (leriglitazone), Minoryx Therapeutics' treatment for cerebral adrenoleukodystrophy in boys aged 2 to 12 with early-stage lesions. The Barcelona company saw the same drug rejected two years ago.

This time the committee took the Phase 2/3 NEXUS1 data alongside real-world evidence from compassionate use, and recommended authorisation under exceptional circumstances. The disease has no approved treatment in the EU and can kill within three to four years of onset.

Neuraxpharm will commercialise it; the Commission decision is due by the end of September. A file that came back and got through on more evidence, rather than by going to a friendlier regulator, is the more useful precedent for the next European biotech.

(GlobeNewswire) (EMA)

Berlin's Pentixapharm Gets Its Phase 3 Money From Existing Shareholders

Pentixapharm completed a €20.4 million capital increase on July 27. Existing shareholders took up 82.5% of the rights offering during the July 7 to 21 window, worth about €16.8 million, with the rest placed separately.

The money funds PANDA, a registration-enabling Phase 3 trial of the gallium-68 PET tracer PentixaFor, which sorts patients with primary hyperaldosteronism into those helped by adrenal surgery and those better treated with drugs. The tracer picked up FDA Fast Track designation this month and the trial starts this half.

Diagnostics rarely raise like therapeutics. A German radiopharma funding a pivotal trial off its own share register is the unglamorous version of European capital working.

(European Biotechnology)

GSK Leaves Stevenage After Fifty Years

GSK said on July 29 that it will close its Stevenage R&D hub and move around 1,000 staff to a new £400 million site on the Cambridge Biomedical Campus. Roughly 1,800 people work at Stevenage today; the site is earmarked for closure in 2029.

The move sits inside £1.9 billion of planned cost cuts, and concentrates research further into the Cambridge, Oxford, and London triangle instead of widening it. Stevenage grew a cell and gene therapy cluster and a startup campus around GSK's presence.

Those tenants are staying, and the catapult and incubator with them. Whether a cluster survives losing its anchor firm is a question the UK will answer in public by 2029.

(Cambridge Independent)

Quantum & Photonics

Algorithmiq Publishes the Benchmark That Could Break Its Own Claim

IBM and Algorithmiq announced on July 30 a quantum advantage demonstration, simulating a heterogeneous quantum material on an IBM Heron processor in a regime classical methods struggle with. Eight months after the problem entered the Quantum Advantage Tracker, the companies say no classical approach has reliably reproduced the results across the full range.

The framework published alongside it matters more than the claim. It uses controlled noise manipulation, cross-device execution, and error mitigation with quantified uncertainty, so that a computation nobody can check classically can still be trusted.

Algorithmiq, founded in Helsinki and headquartered in Milan since May, is releasing its best classical method as an open benchmark. Handing rivals the sharpest tool to attack your own result is how a field stops arguing by press release.

(IBM)

Space

ESA Tries to Salvage the Module NASA Stopped Needing

ESA issued an invitation to tender on July 27 for studies on repurposing Lunar Link, the communications hardware Thales Alenia Space France built for NASA's Gateway station under a €296 million contract in 2021. NASA paused Gateway in March to pursue a different lunar base strategy.

The hardware has finished functional and environmental testing and should be qualified by the end of 2026. ESA will award up to two parallel fixed-price contracts worth €3 million each for a study capped at 12 months, targeting a launch in the last quarter of 2029.

Thales Alenia Space France is a mandatory subcontractor, with at least €750,000 reserved. Europe built the module and then lost the station it was meant for; whether it flies now rests on a study worth a fiftieth of what building it cost.

(SpaceNews)

Materials & Manufacturing

Brussels Opens a €1.5 Billion Loan Window for Battery Cells

The Commission launched the Battery Booster Facility call on July 28, offering up to €1.5 billion in interest-free loans for battery cell manufacturing across the European Economic Area. This is repayable credit, not grant money.

That distinction shapes who applies. Europe's battery problem stopped being chemistry or announced capacity a while ago; it is plants reaching commercial yield, and a loan selects for applicants who expect to run a line.

(Eunews)

One to Watch

Whether Nexperia Buys Into France's Packaging Plant

Tessalia, the advanced packaging joint venture of Foxconn, Radiall, and Thales, laid its foundation stone at Le Barp in Nouvelle-Aquitaine on June 1, with production due at the end of 2029 and €250 million of investment planned by 2033. Bits&Chips argued on July 27 that Nexperia should take a shareholding.

The Dutch have a plan of their own: Nexperia's subsidiary Itec, the research institute TNO, and several backend firms are working on an automated panel-level assembly and test operation.

Europe imports almost all of its advanced packaging, and both efforts point at 2029. Watch whether they converge on one credible site or split the money across two that are each too small to matter.

(Bits&Chips)

End of article