Infineon's DEEPCRAFT™ AI Hub, is a dedicated online platform designed to accelerate intelligent edge development on Infineon's PSOC™ Edge MCU + NPU family.
SAN JOSE, Calif., Oct. 30, 2025 /PRNewswire-PRWeb/ -- Infineon Technologies AG (FSE: IFX / OTCQX: IFNNY) and embedUR systems, a leader in embedded systems and Edge AI solutions, announce the launch of Infineon's DEEPCRAFT™ AI Hub, a dedicated online platform designed to accelerate intelligent edge development on Infineon's PSOC™ Edge MCU + NPU family. The hub, available at deepcraft.infineon.com, provides developers with access to pre-optimized AI models, implementation resources, and benchmark data—developed in collaboration with embedUR systems.
While demand for AI on the edge is growing rapidly, the industry still faces a shortage of expertise in optimizing and deploying AI workloads on resource-efficient embedded devices. The DEEPCRAFT™ AI Hub directly addresses this gap by offering a collection of ready-to-use models that are ready to run on Infineon platforms without requiring modification, or extensive AI deployment expertise.
By removing the complexity of porting models to PSOC™ Edge and by providing a selection of ready-made models as a starting point, the hub enables developers to bridge the skills divide and accelerate the path from concept to production.
Accelerating Intelligence on PSOC™ Edge
Developed in collaboration between embedUR systems and Infineon, the DEEPCRAFT™ AI Hub serves as a central destination for developers building intelligent, low-power applications on Infineon's PSOC™ Edge devices. As part of this collaboration, embedUR has ported and optimized twenty AI models based on open-source architectures from its ModelNova repository for the PSOC™ Edge E84 device, leveraging the Arm® Ethos-U55 NPU and Vela compiler to deliver efficient, high-performance inference at the edge. In addition to these 20 models, the hub also hosts specialized, ready-to-use models, solutions, and tools created by Infineon – specifically for the PSOC ™ Edge platform.
Every model has been benchmarked for performance and energy efficiency, accompanied by execution scripts and metrics for reproducibility. The collection includes popular edge-ready architectures such as YOLOv8-Nano, MoveNet Lightning, MobileNetV2, ArcFace, among others, covering domains from object detection and pose estimation to audio denoising and semantic segmentation.
"Infineon is committed to enabling the next generation of IoT edge devices by bringing low-power, optimized performance and AI to the edge," said Sarah Hemmer, Managing Director at Imagimob (an Infineon Technologies company). "The DEEPCRAFT™ AI Hub's portfolio of models, expanded through trusted partners like embedUR, scales PSOC™ Edge platform enablement and allows developers to rapidly bring Edge AI capability to their devices. We look forward to this ongoing collaboration."
"As AI moves closer to where data is generated, the complexity of deployment has grown dramatically," said John Marconi, VP of Technology at embedUR systems. "Our work with Infineon on the DEEPCRAFT™ AI Hub reflects our shared vision of simplifying this journey – so developers can focus less on integration hurdles and more on building intelligent products that make a real impact."
Engineering Excellence at the Edge
embedUR systems brings over two decades of embedded systems experience and a proven track record in delivering production-ready AI solutions for silicon vendors and OEMs worldwide.
With hard to match expertise in firmware, middleware, SDKs and AI model optimization, embedUR bridges the gap between algorithms and hardware—enabling high-performance, power-efficient intelligence directly on the edge.
embedUR's AI engineering team specializes in porting, optimizing, and validating neural network workloads across a range of architectures, from MCUs and NPUs to unique SOCs that combine CPUs, DSPs, and AI accelerators.
By combining strong domain knowledge in embedded systems with advanced AI development workflows, embedUR helps its partners shorten time-to-market and achieve consistent performance across diverse hardware platforms.
Through its collaborations with leading semiconductor companies like Infineon, embedUR continues to advance the state of Edge AI, delivering solutions that turn silicon capability into deployable intelligence.
About embedUR systems
embedUR systems is a Silicon Valley–based technology company specializing in embedded systems, connectivity, and Edge AI solutions. With over two decades of experience bringing complex embedded products to high-volume production, embedUR provides engineering services and platforms like ModelNova to accelerate AI deployment on low-power devices. Learn more at www.embedur.ai and www.modelnova.ai.
Media Contact
Mark Cowtan, embedUR systems Inc., 1 6505392446, [email protected], https://embedur.ai
SOURCE embedUR systems Inc.
Share this article