The Intelligent Edge brings the processing of AI algorithms and the taking of resulting actions to the device itself. Cloud Services can be defined, containerized, and deployed to one (or many) devices. Being able to run “AI at the Edge” has multiple benefits:
Faster decision making
There are multiple scenarios where a device cannot wait for the round trip to the cloud and back for actions. One example is a self-driving car that needs to make decisions locally in milliseconds.
There are multiple scenarios where a device cannot wait for the round trip to the cloud and back for actions. One example is a self-driving car that needs to make decisions locally in milliseconds.
Offline operation
When AI algorithms are processed locally, the device can operate without an active network connection. This is beneficial in the areas where connectivity is unreliable, or the device is moving in and out of network connection.
When AI algorithms are processed locally, the device can operate without an active network connection. This is beneficial in the areas where connectivity is unreliable, or the device is moving in and out of network connection.
Event based cloud connection
While some sensors (e.g., temperature) generate relatively small amounts of data, other sensors (e.g., cameras) can quickly generate gigabytes. Sending a full video stream to the cloud for analysis can be cumbersome and expensive. Running AI in the device itself enables can result in only important data being sent to the cloud. For example, a smart camera can be configured to send only pictures of recognized objects.
While some sensors (e.g., temperature) generate relatively small amounts of data, other sensors (e.g., cameras) can quickly generate gigabytes. Sending a full video stream to the cloud for analysis can be cumbersome and expensive. Running AI in the device itself enables can result in only important data being sent to the cloud. For example, a smart camera can be configured to send only pictures of recognized objects.
Addressing privacy concerns
Being able to pre-process data can avoid privacy concerns. For example, a camera that counts people in a public venue like a train station doesn’t need to stream the images taken to the cloud but can send only the totals for reporting and trend analysis.
Being able to pre-process data can avoid privacy concerns. For example, a camera that counts people in a public venue like a train station doesn’t need to stream the images taken to the cloud but can send only the totals for reporting and trend analysis.
DataON is partnering with Strategic Online Systems, Inc. (SOS) for enterprise-grade AI at the Edge. Its Strategic Vision solutions are deployed privately on-premises and can utilize existing security assets. By being on-premises, all your data stays with you, response time is fast, and your local servers are secure by design.
NVIDIA Collaborates With Microsoft to Accelerate AI
Microsoft and NVIDIA are working jointly on advancing the adoption of AI from high-performance computing in the cloud to the intelligent edge across a wide variety of industries. Together, Azure Stack HCI with the powerful NVIDIA GPUs and toolsets form a better-together hybrid solution that allows enterprises to connect to their on-premises assets and perform complex AI analytics in real time. Bringing this type of innovation to the edge offers the power of the cloud without sacrificing data sovereignty.
Using NVIDIA DeepStream with NVIDIA GPU integration for efficient run-time execution of complex AI applications, software partners such as Strategic Online Systems have built end-to-end solutions that run AI and machine learning solutions at the edge.
Let’s Get Started
We’re here to help you get started
on your hybrid cloud journey.
[wpforms id=”7726″ title=”false” description=”false”]