Compact Edge AI Systems for Space-Constrained Industrial Infrastructure

IBASE Technology has released the ASB110-PI900, an ultra-compact edge computing system designed to integrate artificial intelligence inference directly into smart kiosks, retail terminals, and industrial equipment. This platform addresses the growing need within the digital infrastructure sector to process data locally, minimizing latency by positioning computational hardware physically closer to sensors, cameras, and connected devices.

Hardware Architecture and Processing Capabilities
The physical footprint of the system measures 140 x 95 x 60 millimeters, making it suitable for deployment in environments where standard industrial PCs cannot fit. The system architecture is built around the PI900F PICO-ITX motherboard, configured with either Intel Core Ultra 7 155U or Ultra 5 125U processors. To support the heavy multitasking and rapid data processing required for edge AI inference, the system accommodates up to 48 gigabytes of DDR5-5600 memory. This combination of processing power and high-bandwidth memory enables the system to handle complex workloads, such as real-time retail analytics and automated equipment monitoring, without relying on cloud-based processing.

Connectivity and Digital Infrastructure Integration
To interface with a wide range of peripheral devices and industrial sensors, the system incorporates comprehensive input and output connectivity. Display capabilities are managed through an HDMI port and a DisplayPort transmitted via USB Type-C, allowing for dual-display configurations. Peripheral and network connections are supported by three USB 3.2 ports and a 2.5-Gigabit Ethernet (2.5GbE) port, ensuring high-speed data transfer between local devices and broader network architectures.

For storage and wireless expansion, the motherboard includes dual M.2 sockets. These interfaces allow system integrators to install NVMe or SATA solid-state drives alongside optional WiFi and Bluetooth modules, providing flexibility for varying deployment requirements.

System Security and Mounting Configurations
Deploying computing hardware in accessible or remote edge environments requires robust physical and hardware-level security. The ASB110-PI900 utilizes a 12-volt direct current (DC) input and integrates a Trusted Platform Module (TPM 2.0). This cryptographic processor provides hardware-based protection, securing sensitive data and cryptographic keys against unauthorized access. To accommodate different equipment layouts across various industries, the system housing supports multiple installation methods, including standard desktop placement, VESA mounting, and optional wall or DIN-rail configurations.

Additional Context
This section details technical specifications and competitive benchmarking not included in the original news release.

The industrial edge computing market features several ultra-compact systems designed for localized artificial intelligence inference. Manufacturers such as Advantech, AAEON, and Axiomtek produce comparable PICO-ITX and nano-ITX form-factor computers utilizing low-power x86 architectures. Benchmarking in this hardware category typically relies on physical dimensions (overall volume), maximum memory bandwidth, and the neural processing capabilities of the integrated chipset.

While standard compact industrial PCs often utilize older-generation processors limited to standard x86 instruction sets, systems integrating current-generation processors—such as the Intel Core Ultra series—leverage integrated Neural Processing Units (NPUs) to accelerate machine learning workloads without requiring a discrete, high-power graphics processing unit. Objective competitive metrics also include the integration of high-bandwidth networking, such as 2.5GbE interfaces, and hardware-level encryption standards like TPM 2.0 to secure localized digital infrastructure against physical and network-based tampering.

Edited by Aishwarya Mambet, Induportals Editor, with AI assistance.

www.ibase.com