INTELLIGENT TECHNOLOGY

DATA CENTRES

. ANT has deployed its secondgeneration photonic processors in a high-performance computing( HPC) environment at the Leibniz Supercomputing Center( LRZ).

This deployment marks a key step toward integrating light-based co-processing into production data centre operations to reduce the escalating energy and performance constraints of AI workloads.

Photonic computing architectures can enable up to 90x lower power consumption per workload and up to 100x greater data centre capacity – driven by higher computational density and increased calculation speed.

Building on its first-generation system at LRZ, Q. ANT’ s Gen 2 Native Processing Units( NPUs) act as photonic AI accelerators, delivering higher computational throughput and improved energy efficiency. Installed via standard PCIe interfaces, the processors integrate into existing HPC systems and operate alongside CPUs and GPUs under AI and scientific simulation workloads.

“ AI is pushing data centre power consumption to unprecedented levels and energy has become a major limiting factor in scaling next-generation computing infrastructure,” said Bob Sorensen, Hyperion Research ' s Senior Vice President for Research.

“ What makes this deployment significant is that it moves photonic co-processing beyond proof-of-concept and into production HPC environments. Demonstrating measurable energy reduction and performance gains under real-world workloads signals that alternative architectures like photonics are becoming a practical path forward for scaling AI infrastructure.”

In benchmark evaluations at LRZ, Q. ANT’ s Gen 2 architecture demonstrated significant improvements over its first-generation NPUs, marking the next milestone in the company’ s aggressive product development roadmap. Results include:

• More than 50x higher throughput of matrix multiplications

• 25x faster inference on a ResNet-18 convolutional neural network

• 6x lower energy consumption for typical workloads

• Enhanced analog units optimised for nonlinear functions, reducing parameter counts and training depth

• Accuracy sufficient to support state-ofthe-art AI applications

Q. ANT’ s photonic NPUs execute mathematical operations directly in the optical domain using Thin-Film Lithium Niobate( TFLN) photonic integrated circuits – eliminating on-chip heat generation and cooling requirements.

Dr. Michael Förtsch, CEO, Q. ANT, said:“ At LRZ, we’ re proving that light-based co-processing can integrate with today’ s infrastructure and deliver measurable efficiency gains under real workloads. This is how AI can continue to scale without scaling its energy footprint.”

LRZ runs large-scale scientific simulations, AI research and dataintensive applications under the most demanding operational standards. The installation enables LRZ to rigorously test photonic co-processing under production conditions, benchmarking performance, precision and energy efficiency within heterogeneous HPC architectures.

“ This deployment highlights the technological progress from the first to the second generation of Q. ANT’ s processors,” said Prof. Dr. Dieter Kranzlmüller, Chairman of the Board of Directors, LRZ.“ Our evaluation is conducted under real production workloads and operational requirements. Photonic co-processing represents a promising approach to addressing the performance and energy challenges increasingly defining modern high-performance computing.”

The Leibniz Supercomputing Centre( LRZ) is a leading European highperformance computing facility serving universities and research institutions across Germany. As a member of Germany’ s Gauss Centre for Supercomputing( GCS), LRZ provides large-scale computing infrastructure supporting scientific research, AI development and dataintensive applications. •

28

INTELLIGENT CIO EUROPE www. intelligentcio. com