The module brings high-performance, rugged AI computing to industrial and robotics applications, enabling real-time decision-making in constrained and harsh environments.
As AI moves closer to the edge, industrial and robotics applications require compact computing modules that can handle heavy workloads, deliver real-time performance, and operate reliably in harsh environments. Graphics, AI inference, and autonomous navigation increasingly demand modules that combine CPU, GPU, and dedicated AI processing in a single rugged package.
ADLINK has introduced its first Intel Core Ultra Series 3 COM Express module, the Express-PTL, aimed at demanding edge AI applications. The module pairs Intel’s latest Arc GPU with NPU 5.0 for dedicated AI acceleration, delivering up to 180 TOPS of combined compute power. This allows complex AI tasks such as object recognition, real-time decision-making, and graphics-intensive processing to run efficiently in a compact, industrial-grade form factor.
The module supports a hybrid CPU setup with high-performance, efficiency, and low-power cores, and up to 128 GB of DDR5 memory with error correction. Industrial-grade versions can operate in temperatures from –40°C to 85°C and include features like time-sensitive networking and functional safety compliance, making them suitable for mission-critical applications in robotics, automation, and industrial AI.
The company is also expanding its portfolio with the COM-HPC-mPTL, SBC35-PTL, and MXE-330, all built on the same CPU, GPU, and AI acceleration backbone. These solutions are designed for compact and rugged deployments, offering flexible connectivity, scalable performance, and up to 180 TOPS of AI computing power.
Key features of the platform include:
- Up to 180 TOPS combined AI compute with CPU, GPU, and NPU
- Hybrid CPU architecture with up to 128 GB DDR5 memory
- Industrial-grade operation from –40°C to 85°C
- Rugged design for edge and mission-critical applications
- Compact COM-HPC Mini form factor for space-constrained deployments
With the platform, the company claims to deliver a high-performance, reliable, and compact platform that brings edge AI computing to industrial robotics, automation, and graphics-heavy applications, simplifying deployment while maximising efficiency and performance.
