Cloud Container Instance

Product Superiority

Directly use container GPUs with a low price, greatly improving the cost efficiency of GPU instances;Calculate the container fee in seconds and adopt the pay-as-you-go billing mode, greatly reducing users' IT investment costs; Provide security isolation capabilities at the virtual machine level, and also enjoy the start-stop speed of containers; Natively support Kubernetes and Docker interfaces, Docker image format, and open source community ecology.

01Serverless architecture
02Second-level billing
03Security and reliability
04Openness and compatibility

01Serverless architecture

Directly use container GPUs with a low price, greatly improving the cost efficiency of GPU instances

02Second-level billing

Calculate the container fee in seconds and adopt the pay-as-you-go billing mode, greatly reducing users' IT investment costs

03Security and reliability

Provide security isolation capabilities at the virtual machine level, and also enjoy the start-stop speed of containers

04Openness and compatibility

Natively support Kubernetes and Docker interfaces, Docker image format, and open source community ecology

Serverless architecture

Directly use container GPUs with a low price, greatly improving the cost efficiency of GPU instances

Second-level billing

Calculate the container fee in seconds and adopt the pay-as-you-go billing mode, greatly reducing users' IT investment costs

Security and reliability

Provide security isolation capabilities at the virtual machine level, and also enjoy the start-stop speed of containers

Openness and compatibility

Natively support Kubernetes and Docker interfaces, Docker image format, and open source community ecology

Product Features

Create GPU container instances of various specifications; Customize container start commands and environment variables; Provide rich and convenient remote debugging and access methods.

One-stop container lifecycle management

Support the creation, modification, query, and release of A100-80g container instances of four specifications: 1 GPU, 2 GPUs, 4 GPUs, and 8 GPUs; and support login to containers through WebShell;
Support mounting parallel file storages

Support mounting high-speed parallel file storages in training scenarios; - all-around container status monitoring
Support container observability

All-around container status monitoring, provide container instance monitoring of various metrics and container event information.

Application Scenarios

Meet various traffic flow fluctuations and AI training and inference application scenarios.

01AI training and inference
02Internet business

AI training and inference

AI training and inference: Support applications such as AI training and inference to run in containerized mode, and provide hardware acceleration capabilities such as high-performance network and storage to help you use resources flexibly on demand;

Cloud Container Instance provides serverless container services, integrates network and storage services, and features high-performance heterogeneous infrastructures (x86 servers, GPU-accelerated servers), with the containers running directly on physical servers.

Internet business

Internet business: Internet business has periodic changes due to daily traffic, for example, during the peak period, resources need to be rapidly expanded, and during the trough period, a large number of idle resources exist, and there are certain demands on cost. In such cases, container instances can well meet the demands of this kind of business for flexible use on demand.

01AI training and inference

02Internet business

AI training and inference

Internet business