THANK YOU FOR SUBSCRIBING
AWS' Latest Updates: Launch of New HPC Cloud Instances
Amazon Web Services (AWS) has broad and deep cloud-based services that enable consumers to build and monitor IoT applications to meet their needs. Amazon FreeRTOS, a real-time microcontroller operating system, runs on gateways and servers on devices such as industrial sensors and lightbulbs and AWS IoT Greengrass. Today, some of the top and social and video streaming companies like Instagram, Netflix, and many other companies rely on AWS to run their websites.
Amazon Web Service has now added two new High Performance Computing (HPC) cloud instances that are capable of reaching millions of cores which support 100 Gbps networking and an MPI communication supporting network interface. A variant of the original P3 instance released as P3dn is built on the basis of NVIDIA’s V100 Tesla GPU that consists of 16GB of high memory bandwidth. The instance also features 100 Gbps network bandwidth local NVMe storage. This instance will make it twice the local memory capacity.
The 8-GPU version has 48 Intel Xeon Skylake cores (96 vCPUs) and 1.8 TB NVMe-based Solid State Drive (SSD). With twice the GPU memory and 1.5 times as many vCPUs as p316 times large instances, consumers can explore more prominent and complex deeper learning algorithms, render 3D images, transcode video in real time, and model financial risks. The instance of the P3dn GPU is ideal for distributed machine learning and high-performance computing. The price of P3DN is not yet announced AWS, while the on-demand price of the previous version, 8-GPU P3 instance was around $24.48 per hour.
The second new instance is the C5n, which is powered by 3.0 GHz Intel Xeon Platinum 8000 series processors like the older version C5 but surpasses it in large memory storage and faster networking capability features. It greatly increases the maximum available performance in the Amazon Web Service computer-intensive instance family. The C5n has 30 percent more memory capacity and four times more processing speed than its older version. High-performance web service, scientific modeling and simulation, batch processing, distributed analysis, machine or deep learning inference, advertising, highly scalable multiplayer gaming, and video encoding are the focus of C5n instances.
AWS began to compete with rivals such as Microsoft Azure using InfiniBand in its HPC cloud infrastructure. For its new launch, AWS has not revealed the underlying network technology; whether it is based on InfiniBand or Ethernet, Amazon appears to have matched Azure’s raw network bandwidth.