Elastic Compute Service (ECS) is compatible with a wide range of Alibaba Cloud services. You can build applications by using only ECS instances, or pair ECS instances with other Alibaba Cloud services to extend your application capabilities.
This topic describes some typical use scenarios of ECS.
New websites usually have low or sporadic traffic. In cases such as these, you can host your websites on low-configuration instances that can provide sufficient performance for basic website components. As your business grows, you can scale up your instance or scale out to more instances at any time to handle traffic spikes.
Some e-commerce applications, especially applications used in flash sales or promotions, may experience brief, sharp fluctuations in traffic. In cases such as these, you can use Auto Scaling to automatically scale your ECS instances based on business forecasts to maintain steady, predictable performance at the lowest possible cost. You can also leverage Server Load Balancer (SLB) to the mix to improve service availability and ensure consistent user experience. For more information, see What is Auto Scaling? and What is SLB?
ECS caters to data analytics scenarios with its big data instance families. Instances in these instance families are ideal for distributed computing workloads based on the Hadoop framework, log processing jobs, and massive-scale data warehouses. These instance families leverage locally attached storage and interconnected internal networking capabilities to provide immense storage capabilities and minimize network latency for Hadoop and Spark clusters. For more information, see Big data instance families.
GPU-accelerated instances leverage GPUs to deliver accelerated computing capabilities through hardware acceleration. These instances are capable of near-real-time rendering and are ideal for graphics-intensive workloads, such as rendering, cloud graphics workstation, and video transcoding workloads.
vgn6i and gn6i instances are powered by Turing-based NVIDIA Tesla T4 GPUs and are designed to deliver top-of-the-line graphics processing capabilities. vgn6i instances also support virtual GPUs (vGPUs), which allows you to virtualize the equipped GPU into multiple GPUs. This way, you can effectively dedicate portions (1/2, 1/4, or 1/8) of the GPU's computing capabilities towards different tasks.
GPU-accelerated instances provide great price performance balance for time-consuming, compute-intensive deep learning workloads. These instances are ideal for performing floating point computing, data pattern matching, and training deep learning models for machine learning applications. For more information, see GPU-accelerated compute-optimized and vGPU-accelerated instance.
ECS offers instances that are ideal for high-performance computing applications in which you want to run parallel and compute-intensive workloads, such as highly complex simulations. These applications include weather forecasting, biopharmaceutical manufacturing, gene sequencing, and graphics processing.
For highly compute-intensive applications such as these, we recommend that you use Elastic High Performance Computing (E-HPC) to manage ECS with other computing resources. E-HPC helps you aggregate and scale computing capabilities with ease, allowing you to build powerful applications that help solve major challenges across multiple fields, including science, engineering, and commerce.
For more information about the use scenarios of ECS, see Alibaba Cloud solutions.