Q.ANT Brings Commercial Photonic Computing to the US — Launching Austin HQ With 30× Energy Efficiency and a New CTO From IBM
AI / AI Tech Trends | 4 min read
Q.ANT (Stuttgart, Germany; the pioneer in commercial photonic computing; formerly a Bosch spin-out; founded 2018), has announced the opening of its US headquarters in Austin, Texas, and the appointment of semiconductor veteran Bruno Spruth as Chief Technology Officer. As US companies commit to invest over $690 billion into AI in 2026 alone, the limitations of silicon processors are becoming a structural barrier — with data centres consuming exponentially more power and generating more heat than existing infrastructure can handle. Q.ANT was founded specifically to solve this problem: it has developed a processor that computes natively in light, executing complex mathematical operations in the optical domain where heat generation is near zero. By opening in Austin, Q.ANT enters one of the world's leading ecosystems for semiconductor research, hyperscale infrastructure, and advanced computing talent — with plans to grow the US team to 20 employees across software development, photonics, and digital system design within six months.
"Q.ANT's US expansion offers hyperscalers, data centres, and innovators the opportunity to explore how they can reduce energy consumption and improve compute performance through our Native Processing Server. Bruno's deep semiconductor experience and record of global execution is vital for our growth in one of the world's most competitive and innovative markets for artificial intelligence."
— Michael Förtsch, Chief Executive Officer, Q.ANT
Native Processing in Light — 30× Energy Efficiency, 50× Speed, No Active Cooling
At the heart of Q.ANT's technology are photonic chips built on a Thin-Film Lithium Niobate (TFLN) platform — produced on Q.ANT's own pilot line through its Stuttgart-based manufacturing partner. Unlike silicon transistors that generate heat as a by-product of electrical charge movement, photonic processors manipulate light to perform mathematical operations directly — with near-zero on-chip heat generation. The performance differential is striking: the Q.ANT Native Processing Server (NPS) delivers up to 30× greater energy efficiency than conventional CMOS technologies, up to 50× faster computation for matrix-vector operations, and the potential to increase data centre capacity by up to 100× — all without complex active cooling systems. The Leibniz Supercomputing Centre's chairman has compared Q.ANT's chip running at 30 watts against Nvidia GPUs at 700–1,000 watts, noting that electricity has become a fundamental roadblock for AI. Q.ANT's throughput is projected to scale from 0.1 GOps in 2024 to 100,000 GOps by 2028 — a million-fold increase in five years. The NPS is delivered as a 19-inch rack-mountable server, integrating with existing CPUs and GPUs via PCIe and C/C++/Python APIs as a plug-in co-processor — making photonic acceleration immediately deployable without disrupting existing data centre infrastructure.
"Photonics is not an incremental step forward — it is a different way to compute entirely. Q.ANT has built something the industry has needed for a long time, and it is time to bring this technology into the US market at scale. With its combination of top-tier technical universities, a mature semiconductor ecosystem, and supportive regulatory environment, Austin is the ideal home for Q.ANT's US headquarters."
— Bruno Spruth, Chief Technology Officer, Q.ANT
Live HPC Deployments, the $80M Series A, and the Path to Local Manufacturing
Bruno Spruth brings 16 years of senior leadership at IBM, culminating in global responsibility for POWER processor development — technology underpinning mission-critical systems and AI workloads worldwide. In 2025, Q.ANT became the first company in the world to deploy a commercial photonic processor in a live production HPC environment — at the Leibniz Supercomputing Centre (LRZ) in Germany, where its processors actively run workloads in climate modelling, medical imaging, and fusion energy research. A four-year partnership with the Jülich Supercomputing Centre (JSC) further extends commercial photonic deployment across Europe's leading HPC infrastructure. The Generation 2 NPU is available to order with customer shipments in the first half of 2026. In April 2026, Q.ANT also appointed Michael Krueger (20+ years at Intel) as VP of Commercials to lead US sales and business development. The US launch follows the close of an $80 million Series A round — from Duquesne Family Office, Hermann Hauser (ARM founder), Cherry Ventures, UVC Ventures, L-Bank, imec.xpand, Verve Ventures, Grazia Equity, EXF Alpha, LEA Partners, Onsight Ventures, and TRUMPF — bringing total Series A funding to over €142 million. Q.ANT is also expanding with plans to localise chip manufacturing in North America and is included in the Gartner Hype Cycle for Data Centre Infrastructure Technologies 2025.
Key Takeaways
- • Q.ANT (Stuttgart, Germany; CEO Michael Förtsch; pioneer in commercial photonic computing; formerly Bosch spin-out; founded 2018) has opened its US headquarters in Austin, Texas, and appointed Bruno Spruth (former IBM VP Power Processor Development; 16 years IBM) as Chief Technology Officer — announced 23 April 2026. US headcount target: 20 employees (software development, photonics, digital system design) within six months. Plans to localise chip manufacturing in North America.
- • The technology: Q.ANT's Native Processing Server (NPS) computes complex mathematical operations natively in light — using photons instead of electrons. Built on a Thin-Film Lithium Niobate (TFLN) platform. Key performance claims vs conventional CMOS: up to 30× greater energy efficiency; up to 50× faster computation (matrix-vector operations); up to 100× data centre capacity increase — all without complex active cooling. Q.ANT chip: 30 watts vs Nvidia GPU 700–1,000 watts. IEA 2026: data centre energy use projected to surpass Japan's entire annual electricity consumption.
- • Deployment and product: NPS is a 19-inch rack-mountable server integrating with existing CPUs and GPUs via PCIe and C/C++/Python APIs — immediately deployable as a plug-in co-processor. In 2025, Q.ANT became the first company in the world to deploy a commercial photonic processor in a live production HPC environment at Leibniz Supercomputing Centre (LRZ), Germany — running climate modelling, medical imaging, and fusion energy research workloads. Four-year partnership with Jülich Supercomputing Centre (JSC). NPU Generation 2 available to order; customer shipments H1 2026.
- • Throughput roadmap: 0.1 GOps (2024) → 100,000 GOps (2028) — a million-fold increase in five years. US market context: US companies commit to invest $690B+ into AI in 2026; silicon processors and accelerators pushing against computation limits; data centres consuming exponentially more power and generating more heat. Q.ANT's photonic approach targets AI inference, physics simulations, image analysis, non-linear mathematical applications, and time-series analysis.
- • Funding and leadership: $80M Series A — Duquesne Family Office (lead), Hermann Hauser (ARM founder), Cherry Ventures, UVC Ventures, L-Bank, imec.xpand, Verve Ventures, Grazia Equity, EXF Alpha, LEA Partners, Onsight Ventures, TRUMPF; total Series A funding €142M+. Also in April 2026: Michael Krueger (20+ years Intel) appointed VP of Commercials for US expansion. Gartner Hype Cycle for Data Centre Infrastructure Technologies 2025 inclusion. Bruno Spruth's framing is definitive: "Photonics is not an incremental step forward — it is a different way to compute entirely."