Demonstrations from iENSO and Hellbender highlight rapid porting, agentic AI at the edge, and next-generation performance on Ambarella's N1-655.
SAN DIEGO and CHATOU, France, March 10, 2026 /PRNewswire-PRWeb/ -- Macnica today announced live demonstrations from two of its Macnica Imaging Ecosystem partners—iENSO and Hellbender—featuring next-generation Edge AI capabilities powered by Ambarella's N1 platform.
The demonstrations highlight agentic AI use cases running at the edge, powered by Ambarella's N1-655 platform, showcasing the ability to run multi-billion-parameter AI models with exceptional performance per watt at low-power envelopes. Together, the demonstrations highlight how Macnica connects Ambarella's silicon platform with ISV innovation to deliver complete Edge AI solutions—from capture and processing to actionable communication at the edge.
Advancing Agentic AI at the Edge
Both iENSO and Hellbender successfully ported AI-driven applications from their ecosystems to the N1-655 within weeks, demonstrating the flexibility of Ambarella's CVflow® architecture and the scalability of the N1 platform for production deployments through its Developer Zone.
iENSO – Industrial Agentic AI on N1
Exhibiting in the Macnica booth (Hall 2/2-647), iENSO will demonstrate an Industrial Agentic AI system running fully on the N1-655 platform, bringing large-model intelligence directly onto the factory floor.
The demo integrates: Hybrid Distributed and Centralized AI Processing
Demonstrates a scalable architecture combining iENSO's Ambarella-based vision platforms (CV28M and CV75) with the Ambarella N1-655 centralized AI compute platform, enabling efficient distributed inference alongside centralized orchestration of AI workloads.
Real-Time Industrial Inspection
Executes production-grade computer vision models for automated product and packaging inspection, delivering reliable real-time analytics suitable for modern manufacturing environments.
AcuSight Industrial AI Platform
Provides a comprehensive software layer for device orchestration, ML lifecycle management, and operational visibility across factory, plant, and enterprise environments. The platform enables real-time sensor data aggregation, KPI dashboards, and centralized device fleet management.
Secure On-Premises Deployment with Full Data Sovereignty
Architected for environments with strict security and regulatory requirements, the system operates entirely at the edge without reliance on cloud connectivity, ensuring that operational data, video streams, and AI insights remain within the customer's infrastructure.
Agentic AI-Assisted Operations
Integrates a local LLM-based agent to support natural language interaction with the system, providing contextual guidance on operational status, procedures, performance queries, and operator assistance.
In addition to N1, iENSO will demonstrate how Ambarella's CV75 and CV28 Edge AI vision SoCs enable scalable architectures for multi-sensor camera streaming and distributed inference. Together, these platforms illustrate a clear migration path from today's edge AI deployments to emerging agentic AI workloads powered by the N1 platform.
Hellbender – Spatial Intelligence & Embedded GenAI
Demonstrating in Ambarella's booth (Hall 5 / 5-355), Hellbender showcases high-performance spatial awareness and on-device Generative AI running entirely on the Ambarella N1-655. The demonstration highlights how a unified architecture scales across applications ranging from roadway safety monitoring to industrial AMR operations.
The demo consolidates three AI tiers onto a single SoC:
- Real-Time Perception: Object detection, optical flow, and depth estimation using CVflow® Neural Vector Processor (NVP)
- Spatial AI: Real-time visual odometry and actor tracking to build digital twins of the environment.
- Agentic AI: On-device Visual Language Models (VLMs) generating contextual scene understanding.
Hellbender provides examples of a unified software stack processing monocular camera data from distinct environments: a roadway safety application and a factory floor setting. For road safety, the SoC acts as an AI-enabled dash camera, generating live 3D spatial maps and performing actor tracking. For industrial applications, the system recreates a mobile robot navigating a factory floor.
Across both applications, the integrated VLM enables natural language queries of the scene. This translates complex 3D sensor data into actionable, natural language insights—creating a more accessible interface for non-technical users in the field.
This architecture demonstrates how the N1-655 platform moves beyond traditional robotics workloads—handling standard perception and spatial reasoning while extending into language-based reasoning for complex scenes—all from a single edge AI device.
Platform-Level Innovation with Rapid Porting
The rapid onboarding of both ISVs underscores the simplicity and scalability of Ambarella's CVflow® AI engine and N1 architecture as well as the ability of the Developer Zone to drive quick design integration.
Within weeks, both partners were able to:
- Port complex AI workloads to production-ready silicon
- Optimize multi-tier perception + GenAI pipelines
- Run multi-billion-parameter models with strong performance-per-watt efficiency
- Validate real-world industrial and mobility use cases
This speed-to-deployment demonstrates the strength of the Macnica Partner Ecosystem and the Ambarella Platform & Developer Zone in accelerating time-to-value for next-generation Edge AI platforms.
Enabling Ambarella's Next Go-To-Market Strategy
Macnica has played a strategic role in bringing both iENSO and Hellbender into Ambarella's ecosystem to jointly deliver solutions tailored for independent software vendors (ISVs). By bridging advanced silicon capabilities with real-world application requirements, the collaboration enables ISVs to build and deploy high-performance edge AI solutions more rapidly and efficiently.
These demonstrations support Ambarella's evolving N1 go-to-market strategy and reinforce Macnica's role in enabling full "Capture → Process → Communicate" architectures—from components to complete Edge AI solutions.
About Macnica ATD Europe
Macnica is a trusted technology solutions partner, empowering the world's leading businesses to succeed in the future. Macnica's products, services, and innovations are crafted to future-proof organizations by continually pushing the boundaries of what is possible.
Bold collaboration with partners and customers drives their products and services to new levels of achievement. By accelerating digital transformation, Macnica helps ignite exponential growth in technology, delivering forward-looking value that influences businesses, economies, society, and the planet.
Together with global partners, Macnica leverages the latest advancements in semiconductors, imaging, secure networks, and artificial intelligence to address tomorrow's challenges.
Macnica ATD Europe, a subsidiary of Macnica Inc., operates across Europe, with offices in France (Chatou), Germany (Munich, Ingolstadt, Regensburg, Bernburg), and the UK (Milton Keynes), providing expertise and innovative solutions to this dynamic region.
For more information, please visit www.macnica.com/eu
About Ambarella
Ambarella's products are used in a wide variety of edge AI and human vision applications, including video security, advanced driver assistance systems (ADAS), electronic mirrors, telematics, driver/cabin monitoring, autonomous driving, edge infrastructure, drones and other robotics applications. Ambarella's low-power systems-on-chip (SoCs) offer high-resolution video compression, advanced image and radar processing, and powerful deep neural network processing to enable intelligent perception, sensor fusion and planning.
For more information, please visit www.ambarella.com.
Media Contact
Carl Dahlberg, Macnica, 1 5129474662, [email protected], https://www.macnica.com/eu/atd-europe/en/
SOURCE Macnica
Share this article