Before you use Machine Learning Platform for AI (PAI), we recommend that you activate PAI and purchase resources for AI development and training. To use cloud-native resources, you must purchase the resources, create a resource quota, and associate the resource quota with a workspace. To use big data engine resources, you need only to purchase the resources and associate the resources with a workspace. This topic describes cloud-native resources and big data engine resources.
Resource types
Cloud-native resources
Lingjun resources
Lingjun resources are computing resources provided by PAI for large-scale deep learning and integrated intelligent computing. Based on a seamless blend of software and hardware enhancements, Lingjun resources deliver the high-performance and heterogeneous computing power that is essential for high-performance computing. Lingjun resources provide core benefits such as high performance, high efficiency, and high resource utilization. You can use Lingjun resources for AI development, training, and inference in PAI.
General computing resources
General computing resources, such as Elastic Compute Service (ECS), Elastic Container Instance, and Elastic GPU Service (EGS) resources, provide flexible, stable, easy-to-use, and high-performance environments for training deep learning models. After you activate PAI, the system automatically creates a public quota for general computing resources and provides heterogeneous resources of multiple types. You can associate the resource quota with a workspace to use the resources.
Big data engine resources
MaxCompute resources
MaxCompute is an enterprise-level cloud data warehouse that uses the software as a service (SaaS) model. MaxCompute is suitable for scenarios that involve data analytics. It provides a fast, fully managed online data warehousing service in a serverless architecture. MaxCompute eliminates the constraints of traditional data platforms in terms of resource extensibility and elasticity, minimizes O&M costs, and allows you to efficiently process and analyze large amounts of data at low costs. For more information about MaxCompute resources, see What is MaxCompute? You can purchase different types of MaxCompute resources, create a resource quota, associate the quota with PAI to develop models, and then submit the models to MaxCompute for large-scale distributed training.
Fully managed Flink resources
Realtime Compute for Apache Flink is a one-stop real-time big data analytics platform developed by Alibaba Cloud based on Apache Flink. Realtime Compute for Apache Flink supports end-to-end data analytics within subseconds. For more information, see What is Alibaba Cloud Realtime Compute for Apache Flink? You can purchase fully managed Flink resources on the Resource Quota page in the PAI console and associate the resources with a workspace to perform model training.
Guidance
Log on to the PAI console. In the left-side navigation pane, choose AI Computing Resources > Resource Pool. On the Resource Pool page, create resource groups of different types and purchase computing resources.
For more information about how to purchase and manage general computing resources, see Create and manage general computing resources.
For more information about how to purchase and manage Lingjun resources, see Use and manage Lingjun resources.
Log on to the PAI console. In the left-side navigation pane, choose AI Computing Resources > Resource Quota. On the Resource Quota page, create resource quotas for different types of resources. You can associate the resource quotas with a workspace to perform AI development and training.
Cloud-native resource quotas
After you purchase cloud-native resources on the Resource Pool page in the PAI console, you can allocate the computing resources in one or more resource groups to one or more resource quotas. You can create child-level resource quotas to generate a quota tree that enhances performance in queuing and scheduling.
For more information about how to create and use a resource quota for Lingjun resources, see Lingjun resource quotas.
For more information about how to create and use a resource quota for general computing resources, see General computing resource quotas.
Big data resource quotas
For more information about how to activate MaxCompute, purchase MaxCompute resources, and create and use a resource quota for MaxCompute resources, see MaxCompute resources.
For more information about how to activate Realtime Compute for Apache Flink, purchase fully managed Flink resources, and create and use a resource quota for the resources, see Fully managed Flink resources.