Tag

EDGE COMPUTING

Browsing

The American chipset manufacturer Nvidia is rolling out its EGX platform to bring real-time artificial intelligence (AI) to the edge of the network. Nvidia EGX is an accelerated computing platform, which boasts low-latency AI that enables companies to perceive, understand and act in real time on continuous streaming data between 5G base stations, warehouses, retail stores, factories and beyond. AI computing will occur at the edge of the network, where sensors collect data before it is sent to cloud-connected data centers. This platform is expected to satisfy the increasing demand for AI applications to assist numerous devices streaming continuous raw sensor data and is designed for high-throughput AI at the edge where data is primarily sourced to achieve instantaneous and guaranteed response times while reducing bandwidth to the cloud. According to the company, by 2025, there could be up to 150 billion machine sensors and the Internet of Things (IoT) devices streaming…

Verizon has announced a successful edge computing trial on a live 5G network, conducted using their newly formed 5G test bed in Houston. The company’s engineers used Multi-Access Edge Compute (MEC) equipment and MEC platform software, incorporated in a network facility closer to the network edge, and managed to decrease network latency by half. The test results are important for next generation networks, enabling 5G to support low-latency applications such as Virtual Reality (VR). Low latency refers to minimizing the round-trip time taken by data packets. This is an essential yet enormously challenging attribute for a 5G network, and its high performance requirements. At present, many mobile applications depend on cloud-based processing that is handled in centralized data centers, and the physical distances involved increase latency. According to Verizon, locating the computing power nearer to the users at the network edge significantly decreases the time to deliver services. …