Integrating AI-ML at the Edge with the SOSA Technical-Standard

September 16, 2022

AI continues to have its impact on military computing projects, and applications involving the new architecture SOSA is no exception.

 

This webcast will examine how the newly released Sensor Open Systems Architecture (SOSA) leverages AI technologies to implement the use of video and data analytics at the edge with building blocks aligned to the open standard.

 

Key building blocks will be introduced using a mock proposal and sample payload. Through this exercise, the speakers will demonstrate:

 

How to build a system
Which elements to include
How to scale the technologies
Ease of system configuration using SOSA
The benefits of interoperability
Technical details covered will include:

 

How to select a backplane to build the overarching reference architecture
NVIDIA's Ampere architecture, which provides dedicated cores for AI optimization and addresses low latency requirements critical to military applications
The latest single board computers using Intel Tiger Lake chipset supporting PCIe Gen 4, and how the heart of the system interconnects the components
Addressing security questions with separate data and control planes
How data gets passed and processed within the system GPGPU

Related Blog Post

Elma's Wistia Channel

Watch Here

Watch More Webinars

https://elma-2.wistia.com/medias/my27eywk83

What is Coming Next in 100Gb High-Speed Ethernet Connectivity?

November 22, 2023

The widespread adoption and successful early implementations of SOSA are gaining momentum both in the United States and globally and is increasingly seen in programs of record. This webinar delve into upcoming developments that have the potential to elevate your designs to new levels of performance.

https://elma-2.wistia.com/medias/ihetr599k2

How to Apply Open Standards to Systems Using Radio Frequency Architectures

September 26, 2023

This webinar pulls together experts in design and implementation of MOSA compliant hardware and software components that can be used to develop and deploy RF designs in field applications