Basic Concepts

Terms

Introduction

GPU

Graphics Processing Unit (GPU). Compared with CPU, GPU has many computing units and more pipelines and is suitable for large-scale parallel computing.

CUDA

NVIDIA's universal parallel computing architecture to help you solve complex computing problems with NVIDIA GPUs.

cuDNN

NVIDIA's GPU-accelerated library for deep neural networks.

Image

Provides the information needed to run an instance, including the operating system, initialized application data, and so on.

Region and availability zone

Physical locations where instances and other resources are deployed.

SSH key pair

A secure and convenient login authentication method that consists of a public key and a private key. Only Linux-based instances are supported.

Security Group

A virtual firewall that allows you to control inbound and outbound traffic of instances based on security groups.

ENI

An independent virtual NIC that can be bound to or unbound from an elastic cloud server for flexible application expansion and migration.

EVS

An elastic block storage device that can be used as a system disk or an expandable data disk for an instance.

Snapshot

A backup file of the EVS data status at a point in time, used to back up or restore the entire EVS.

VPC

Virtual Private Cloud. A VPC provides an isolated and network environment for cloud resources such as Elastic Cloud Server (ECS) instances, cloud containers, and cloud databases. VPC's ample features allow you to flexibly manage your cloud networks, including creating subnets, configuring security groups and network ACLs, managing route tables, and applying for elastic IP addresses (EIPs) and bandwidth.