White Paper

White Paper: Accelerating Big Data Processing and AI at the Edge

Issue link: https://read.uberflip.com/i/1409882

Contents of this Issue


Page 1 of 6

WHITE PAPER Accelerating Big Data Processing and AI at the Edge mrcy.com 2 mrcy.com 2 AI AT THE EDGE IS THE NEW REALITY Ar tificial Intelligence (AI) applications powered by deployable GPUs are moving to the edge, adjacent to new generations of advanced sensors and connected by extremely high-bandwidth 5G networks. The big data created by these sensors will not over whelm this new style of systems architecture; this architecture will be used to enable a new reality of edge capability. Data from imaging sensors will be manipulated by AI capabilities to deliver advanced digital maps, image recognition, object tracking and image correction. In some cases, the imaging data will be fused with data from other sensors to enable VR or AR headset displays. Autonomous vehicles will use AI not just for safe movement, but for flexible mission execution. As hundreds of other applications at the edge emerge—built on power ful GPUs, 5G communications and AI—program managers, system architects and design engineers will find new and innovative ways to employ AI, providing capabilities not yet imagined. 5G NETWORKS NEED EDGE INTELLIGENCE Emerging 5G networks are enabling a new type of distributed systems architecture based on edge computing nodes that communicate with each other, and with computing cloud centers, via low-latency high-bandwidth links. In mission-critical systems, these advanced edge nodes will be embedded in a variety of mobile platforms and will suppor t decision-making in tactical command centers and life-saving surgeries in mobile hospitals. The per vasive 5G-connected devices will generate zettabytes of captured information, volumes of big data so large that only a fraction can be realistically transferred to cloud data centers, even at 5G bandwidths. To be effective, new systems architectures must per form analytics and decision-making in these edge nodes, where data transfers through communication links when needed and nodes function independently, using local data whenever possible. Powerful sensing, computing and communications technologies are maturing together, enabling revolutionary changes for aerospace and mission-critical operations. Intelligent edge nodes, collecting vast amounts of data, will soon be communicating via 5G networks with immense bandwidths. System designers are now pulling these capabilities together, using them to drive sophisticated new applications that deliver higher efficiencies, faster responses and better decisions. AI at the Edge Processing at 5G speeds next to the data source enables artificial intelligence for real-time decision-making.

Articles in this issue

Links on this page

view archives of White Paper - White Paper: Accelerating Big Data Processing and AI at the Edge