Q.ANT said has invested € 14 million in machinery and equipment to complement the new line for photonic chips.
“This innovative manufacturing approach delivers faster, more energy-efficient processors to meet the growing computational demands of AI and HPC, while also establishing a blueprint for cost-effectively modernizing chip production worldwide,” the companies said in their announcement. “This groundbreaking initiative establishes a blueprint for democratizing production capacity. It enables countries to attain greater semiconductor manufacturing resilience, reduce dependency on global supply chains, and accelerate the development of critical technologies that drive innovation across data centers, research institutions, and advanced industries.”
The official launch event was attended by industry figures and German officials. By modernizing existing chip production capabilities, Q.ANT and IMS CHIPS say they have pioneered a scalable approach to bringing energy-efficient AI processors to the market at a faster pace, more cost-effectively, and more sustainably.
Q.ANT’s photonic chips – which compute using light instead of electricity – deliver a 30-fold increase in energy efficiency and a 50-fold boost in computing speed, offering transformative potential for AI-driven data centers and HPC environments.
The pilot line is designed for production using TFLN, the optimal material for photonic computing and critical to the success of the technology. TFLN enables ultra-fast optical signal manipulation at several GHz without the need for heat to modulate the light on the photonic circuit. This advantage leads to more precise and energy-efficient control of the light, resulting in a significant increase in computing power and energy efficiency compared to traditional silicon.
“This approach establishes a new benchmark for AI chip manufacturing, providing a path towards greater self-sufficiency and more energy-efficient chip solutions,” said Dr. Michael Förtsch, CEO of Q.ANT. “As AI and data-intensive applications push conventional semiconductor technology to its limits, we need to rethink the way we approach computing at the core. Q.ANT is driving this shift with photonic computing to achieve unprecedented energy efficiency and computational density. With this pilot line, we are accelerating time to market and laying the foundation for photonic processors to become standard coprocessors in high-performance computing. This milestone marks a major step toward the future of sustainable AI chip technology, engineered and produced in Germany for a rapidly evolving global market. By 2030, we aim to make our photonic processors a scalable, energy-efficient cornerstone of AI infrastructure.”
“This pilot line at IMS CHIPS demonstrates how transformative technologies can thrive on existing infrastructure, setting a blueprint for energy-efficient next-generation computing,” said Prof. Dr. Jens Anders, Director and CEO of IMS CHIPS. “This comes at a critical time for the computing industry, as the exponential growth of AI and data-intensive applications will soon overwhelm the current data center infrastructure. By partnering with Q.ANT, we are leveraging our semiconductor manufacturing expertise to accelerate the industrialization of photonic processors and establish a scalable model for energy-efficient computing – a crucial step for the future of AI.”
“Six years ago, we made a bold bet on thin-film lithium niobate, and today that decision gives us a significant advantage,” said Förtsch. “By combining our photonics expertise with our end-to-end control of the value chain – from raw material to finished processor – we are uniquely positioned to drive the next generation of computing and reshape the power and performance challenges of AI and HPC.”
Q.ANT’s photonic approach harnesses light instead of electrons, offering a paradigm shift in computing efficiency and enabling faster and more energy-efficient mathematical operations compared to traditional CMOS processors. Q.ANT has already demonstrated the potential of the technology in cloud-accessible AI inference demos. With PCIe integration, Q.ANT’s Native Processing Servers can seamlessly integrate into existing HPC servers, accelerating adoption across industries.
By leveraging photonics, Q.ANT’s Native Processing Servers can accelerate key workloads such as:
- AI model training and inference
- Scientific and engineering simulations
- Real-time processing of complex mathematical equations
- High-density tensor operations for machine learning
“We are not replacing GPUs – we are reshaping the next generation compute ecosystem,” Dr. Förtsch summarized. “Just as GPUs have complemented CPUs, photonics will enable the next leap in AI – sustainably.”
