IBASE Launches High-Performance Edge AI Single Board Computer

The platform is designed around a heterogeneous computing architecture that combines central processing units, graphics processing units, and neural processing units on a single board. Available processor options include the Intel Core Ultra X7 358H, Intel Core Ultra 7 356H, and Intel Core Ultra 7 355. This arrangement allows workloads to be distributed across specific processing engines to accelerate local inference while reducing network latency and cloud dependency.

In its maximum configuration, the system delivers up to 180 TOPS of total platform AI performance. Memory support consists of two DDR5-7200 SO-DIMM sockets capable of accommodating up to 128GB of system memory to sustain data-intensive tasks and concurrent operations.

Industrial Connectivity and System Expansion
Input and output options comprise dual 2.5GbE LAN ports, USB 3.2 ports, USB 2.0 ports, and two RS232/422/485 serial interfaces. Visual output is supported through HDMI, DisplayPort, USB Type-C with DisplayPort pass-through, LVDS, and eDP connections.

Expansion capabilities include three M.2 slots designated for high-speed NVMe storage drives and wireless communication modules. Power input ranges from 12V to 24V DC, and system security and manageability are reinforced through Intel AMT, a discrete TPM, digital I/O channels, and a programmable Watchdog Timer.

Target Application Sectors
The hardware platform is engineered for deployment in machine vision systems, smart manufacturing lines, intelligent transportation networks, medical imaging equipment, and smart retail terminals. These environments benefit from localized data processing capabilities, which maintain system responsiveness independent of external network availability.

Additional Context:
This section details technical specifications and competitive benchmarking not included in the original product announcement

The IBASE IB966 competes within the compact industrial single board computer market, where form factor constraints must be balanced against rising thermal design power and compute density requirements. Comparable 3.5-inch and small-form-factor embedded boards utilizing mobile processor architectures typically scale up to 128GB of DDR5 memory, but older generations lacked integrated heterogeneous engines capable of surpassing 50 total platform TOPS.

By leveraging Intel Core Ultra Series 3 silicon built on advanced sub-nanometer process nodes, the platform achieves higher transistor density and lower active power consumption per watt than preceding architectures. Competitive benchmarking indicates that systems integrating multi-engine NPU and GPU acceleration exhibit lower execution latency for computer vision and local neural network inference compared to CPU-only or discrete accelerator implementations of equivalent physical volume.

Edited by Sucithra Mani, Induportals editor – adapted by AI.

www.ibase.com