Unlocking Edge AI Vision: VEST SMTX i.MX95-D Development Kit with DeepX

Unlocking Edge AI Vision: VEST SMTX i.MX95-D Development Kit with DeepX (DX-M1)

The VEST SMTX i.MX95-D Development Kit is a high-performance solution purpose-built for Edge AI Vision applications. At its core, the platform features a VEST SMX i.MX95 System on Module (SOM) paired with a robust mini-ITX carrier board that hosts the integrated DeepX (DX-M1) AI accelerator.

By combining the low-latency 10GbE connectivity with support for both USB and MIPI-CSI camera interfaces, this kit unlocks the full potential of NXP’s and DeepX’s processors for real-time intelligence at the edge.

Engineered for Efficiency & Performance

Key architecture features include:

  • Six Arm® Cortex®-A55 cores for high-level application processing
  • Independent Cortex®-M33 and Cortex®-M7 real-time domains
    for deterministic control and safety-oriented design
  • Arm® Mali™-G310 GPU supporting industrial-standard OpenGL® ES 3.2, Vulkan®3, and
    OpenCL® 3.0, which enables high-fidelity 3D graphics rendering and responsive
    dashboards for advanced HMI applications

Image: VEST SMX i.MX95 SOM

Advanced Imaging Pipeline

Optimized for demanding image requirements, the kit features an ISP capable of 500 Megapixels/s throughput:

  • Support high-resolution imaging up to 12MP at 45fps
  • Integrated RGB+IR and HDR capabilities for superior clarity in variable lighting conditions
  • A versatile VPU for 4K video encoding and decoding (264/H.265),
    ideal for high-speed robotics and automated visual inspection applications

Dual-NPU Architecture

At the forefront of Edge AI, the kit leverages a system level Dual-NPU architecture for simultaneous multi-model operations:

  • NXP eIQ® Neutron NPU (2 TOPS): Optimized for low-latency, front-end tasks
  • DeepX (DX-M1) AI Accelerator (25 TOPS): Engineered for high-complexity inference tasks

This high‑performance AI compute capability enables rapid decision‑making for public safety, smart cities, and Industry 4.0 applications. With comprehensive support for TensorFlow, PyTorch, and ONNX via the NXP eIQ® and DeepX DXNN® SDKs, the VEST SMTX i.MX95-D Development Kit ensures that AI deployment is faster, more scalable, and highly power-efficient across medical, retail, and manufacturing sectors.

Image: DeepX (DX-M1) M.2 AI Accelerator Module

Use Case: Automated Quality Inspection (AQI)

In a production line environment, the kit can be deployed:

  • Capturing high-quality images via MIPI-CSI or USB cameras
  • Processing low-latency front-end tasks (image enhancement/ anomaly pre-screening)
    via the integrated NPU (NXP Neutron NPU)
  • Accelerating compute-intensive defect-detection in real time with the DeepX (DX-M1)
  • The Arm® Mali™‑G310 GPU visualizes results through a responsive dashboard with live video, pass/fail indicators, defect highlights, and operator alerts
  • Streaming & Archiving H.264/H.265 inspection video over 10GbE to backend systems such as MES or analytics platforms, along with inspection results and traceability data

Image: Applications

For more information on VEST SMTX i.MX95-D Development Kit with DeepX (DX‑M1), please visit our products or reach out to our sales team at sales@apc-vest.com.