APEX Series Edge AI Platforms
- Technology
- Embedded AI technology
- Partner
- Innodisk
The Innodisk APEX series consists of rugged, industrial-grade Edge AI computing platforms built to deliver reliable on-premises AI training and inference in harsh and demanding environments. Designed for industrial deployment, these systems combine high-performance CPUs with advanced AI acceleration technologies from NVIDIA, Intel, and Qualcomm, enabling fast, efficient, and secure AI processing directly at the edge.
By bringing artificial intelligence closer to the source of data, the APEX series helps reduce latency, improve real-time decision-making, and minimise reliance on cloud connectivity. This makes it ideal for applications such as machine vision, smart manufacturing, automation, robotics, transportation, and other mission-critical industrial workloads where durability and performance are essential.
Range features
A high level overview of what this range offers
-
Rugged design – Designed for tough environments with wide operating temperature support, along with proven shock and vibration resistance
-
Industrial-grade components – Uses Innodisk’s own industrial-grade DRAM and SSDs to ensure durability, endurance, and consistent performance
-
On-premises AI – Enables local AI training and inference to reduce cloud dependency while improving data privacy, latency, and regulatory compliance
-
Versatile I/O connectivity – Provides flexible expansion and connectivity options, including high-speed Ethernet and specialised interfaces such as MIPI over Type-C for camera integration
-
Remote management – Supports Out-of-Band (OOB) remote management, allowing system monitoring and control even if the operating system becomes unresponsive
-
Software support – Compatible with Linux Ubuntu and Windows IoT, and supports AI development tools such as Intel OpenVINO for faster deployment and optimisation
Downloads
for APEX Series Edge AI Platforms
What’s in this range?
All the variants in the range and a comparison of what they offer
| Model | AI / Compute Hardware | Key Performance | Best For / Use Cases | Rugged / Industrial Design | On-Prem AI | I/O + Connectivity | Remote Mgmt (OOB) | OS + AI Software |
|---|---|---|---|---|---|---|---|---|
APEX-X100 | NVIDIA RTX 6000 Ada GPU | 18,176 CUDA / 568 Tensor | On-prem LLM training/inference, HPC | Wide temp support, shock & vibration tested | Local training + inference | High-speed Ethernet, industrial I/O | Out-of-Band remote management | Ubuntu, Windows IoT, AI tool support |
APEX-X100-Q | Qualcomm Cloud AI 100 Ultra | 870 TOPS (INT8) | High-throughput inference, GenAI | Wide temp support, shock & vibration tested | Local training + inference | High-speed Ethernet, industrial I/O | Out-of-Band remote management | Ubuntu, Windows IoT, AI tool support |
APEX-P100 | NVIDIA RTX 5000 Ada (MXM Type B) | 9,728 CUDA / 304 Tensor | Factory AOI, medical imaging | Wide temp support, shock & vibration tested | Local training + inference | High-speed Ethernet, industrial I/O | Out-of-Band remote management | Ubuntu, Windows IoT, AI tool support |
APEX-P200 | NVIDIA RTX 2000 Ada (MXM Type A) | 3,072 CUDA / 96 Tensor | Compact edge AI, wide-temp apps | Wide temp support, shock & vibration tested | Local training + inference | High-speed Ethernet, industrial I/O | Out-of-Band remote management | Ubuntu, Windows IoT, AI tool support |
APEX-E100 | Intel Core Ultra CPU + NPU | Up to 36 TOPS | Edge vision, smart city | Wide temp support, shock & vibration tested | Local training + inference | High-speed Ethernet, MIPI over Type-C camera support | Out-of-Band remote management | Ubuntu, Windows IoT, Intel OpenVINO support |
APEX-S100 | Dual double-width NVIDIA GPUs (up to RTX 6000 Ada) | Multi-GPU scalable | Large-scale video analytics, rack AI | Wide temp support, shock & vibration tested | Local training + inference | High-speed Ethernet, industrial I/O | Out-of-Band remote management | Ubuntu, Windows IoT, AI tool support |
