Rugged embedded computing system for machine learning and artificial intelligence (AI) introduced by Elma

FREMONT, Calif. – Elma Electronic Inc. in Fremont, Calif., is introducing the JetSys-5330 rugged small-form-factor (SFF) embedded computing system for rugged deployed deep-learning and artificial intelligence (AI) applications at the edge of the battlefield.

Based on the NVIDIA Jetson AGX Orin system-on-module (SoM), the rugged AI platform delivers computing power of 275 tera operations per second and expansion sites for custom I/O and storage requirements.

By enabling advanced data processing in harsh environments, the JetSys-5330 is for rugged and mission-critical applications that use wireless multi-access edge computing as well as those in remote or outdoor environments. The JetSys-5330 also is for rugged industrial applications like robotics, autonomous mining, agriculture, and construction vehicles.

The NVIDIA Jetson AGX Orin has a 2048-core NVIDIA Ampere architecture graphics processing unit (GPU) with 64 Tensor Cores and a 12-core Arm Cortex-A78AE. Also featured are deep learning and vision accelerators, a video encoder/decoder, high-speed I/O, 204 gigabytes per second of memory bandwidth, and as much as 64 gigabytes of DRAM.

Related: The next generation of open-systems embedded computing standards

The JetSys-5330 combines the NVIDIA Jetson AGX Orin module with Elma's JetKit-3110 carrier card, which provides the necessary power and interfaces for the SoM, as well as three mini-PCI Express slots, and two M.2 slots for custom I/O or storage expansion.

The rugged embedded computing system has a rugged design is IP67 rated and qualified to standards including MIL-STD-810G, MIL-HDBK-704F, MIL-STD-1275D, and others.

The computer's power consumption ranges from 53 to 130 Watts depending on the system configuration. Maximum weight is 16 pounds with an optionally available fan kit.

For more information contact Elma Electronic online at www.elma.com.