On the surface, aviation manufacturers, shopping centers, universities, police departments, and car manufacturers do not seem to have much in common. In fact, they all use and manage hundreds or thousands of cameras within their respective scopes.

In addition, they also use Network Optix. Network Optix is a global software development company headquartered in California, specializing in developing platforms to help various industries manage, record, and analyze massive amounts of video data. Network Optix collaborates with Intel, based on Intel ®  CoRE ™  The Ultra 200H (codenamed Arrow Lake H) series processor has been optimized for its software. Arrow Lake H has a powerful central processing unit (CPU), integrated neural processing unit (NPU), and graphics processing unit (GPU), which can excellently handle complex edge AI.

IP、 Internet protocols and cameras transmit and receive data through the computer network, which is also a typical monitoring that you will see around shopping malls or on university campuses. They transmit data to a central database, so that security personnel or operations centers can monitor it.

AI is changing the way businesses and organizations view and use video data, and its applications extend far beyond the security field. With the help of AI, cameras on manufacturing assembly lines can count the quantity of products and inform companies of how many products are produced each day and when product packaging is done. AI models can alert managers to areas where production may slow down, or when and how problems may arise. And AI camera vision systems aimed at health and safety can issue alerts when employees are not wearing appropriate protective equipment.

Edge processing, refactoring efficiency

The software of Network Optix can process all video data in the cloud, but its true advantage and speed are reflected on the local server side - thanks to the powerful and excellent performance of the Intel Core Ultra 200H processor designed specifically for edge AI applications. Unlike large data centers equipped with dedicated AI infrastructure, edge AI deployment must be seamlessly integrated into space constrained, low-power, and cost sensitive native IT systems. These systems not only handle AI tasks, but also undertake computing, video, and graphics workloads.

James Cox, Vice President of Business Development at Network Optix, said, "Customers can retrieve annual data in seconds. The pain point for customers in the past was that they needed to spend a lot of time loading and reviewing archives, and the old system was difficult to quickly aggregate global data for multinational companies

The technology of Network Optix is highly scalable and capable of processing massive amounts of camera and video data. At present, the company's customers manage a total of 4.5 million cameras worldwide through its system.

Cox explained: "Putting AI on the cloud or other non local environments means that a large amount of data needs to be continuously transmitted to the cloud - about 5MB per second for each camera. When the scale reaches 100 cameras, the amount of data continuously uploaded will surge to 500MB, while thousands of cameras are deployed in many places. Therefore, it is not realistic to use non local AI. edge computing can convert video into pure data, so as to achieve comprehensive monitoring and real-time alarm. ”

Just like aerospace manufacturers, they are important customers of Network Optix. Imagine a rocket launch scenario: a large number of cameras need to be deployed to continuously transmit real-time images of various parts of the rocket to engineers and control centers. And these real-time information are the key to successful launch.