Overview

Batch Compute is a cost-effective and easy-to-use computing service for enterprises and research institutes engaged in big data computing. It intelligently manages jobs and schedules the optimal resources necessary based on the configured batch size, allowing you to focus on analyzing and processing data results.

Benefits
Fully Managed Service

Batch Compute can fully manage resource and process scheduling for batch computing, so you only need to pay attention to the batch computing status once jobs are defined and submitted. Batch Compute takes care of the creation and termination of the CVMs used for computing, deployment and execution of the computing environment as well as storage and management of the raw data and results.

Cost Optimization

Batch Compute can dynamically allocate and elastically scale the computing resources created for your batch tasks based on the number of tasks and resource requirements, reducing the costs of resource usage. Additionally, Batch Compute requires zero upfront costs, allowing you to make full use of resources for computing jobs with even lower overhead.

Powerful Features and Ease of Use

Batch Compute comes with suite of complementary features that help you define and submit computing jobs, such as defining a variety of implementation methods, orchestrating the computing process and monitoring running status. In addition, it can work in tandem with other relevant Tencent Cloud services such as COS. Batch Compute provides one-stop closed-loop services for computing material acquisition, computing job submission and execution and computing result upload.

Features
Fine Task Definition

After you specify the specs of the CVMs used to process the batch tasks, images and code or commands to be executed, Batch Compute will schedule resources and start computing tasks based on the configuration.

Defining CVM Specs

Intended for specifying the CPU, memory, disk capacity and other information of the cloud server.

Defining Executing Image

Intended for specifying the execution environment on which the batch tasks depend; CVM images and Docker can be used to specify the execution environment.

Running Command Lines and Parameters

After the CVMs are created and images deployed, Batch Compute will execute the user-specified command lines and add the user-configured parameters.

Storage Mounting

Batch Compute supports local mounting of common cloud storage services, offering an identical operating experience for files in the cloud and local files and ensuring access speed of each node in case of massive amounts of concurrent access requests to achieve quick file access. For example, when mounting a COS instance, after the address of the instance to be mounted in the batch task definition and the local address are configured, Batch Compute will automatically mount the instance after the computing node is launched.

Task Dependency Modeling

Batch Compute can define dependencies between different tasks and supports launching multiple instances in one job to complete the computation. For example, if one task can be divided into two different phases with each phase requiring different resources, Batch Compute's task dependency feature can help you create two tasks with different resource requirements where the second task depends on the first one.

DAG Process Editing

Dependencies between multiple batch tasks can be set through the DAG syntax to combine multiple tasks into one single batch job.

Multi-instance Concurrence

The number of instances can be specified in the task configuration, and different instances can obtain their own instance serial number using environment variables and work together to complete the same large analytical job.

Scenarios

Business Type

DNA sequencing, drug screening

Suggestions for Cloudification

High-performance computing environments are necessary for enabling labs or genetic companies to quickly obtain the analytical results of high volumes of data and the scientific insights that facilitate product launches. Batch Compute can be readily used in applications such as computational chemistry, clinical modeling, genome sequencing and analytics across an organization.

Business Types

Movie, television and design sketch rendering

Suggestions for Cloudification

Media and entertainment companies need highly scalable batch computing resources to accelerate and automate data processing. Batch Compute gives users the ability to automate content rendering workflows. Users can build their own rendering-dependent processes, while leveraging Batch Compute's abilities to schedule massive volumes of resources and jobs to efficiently complete visual creation work.

Pricing

Batch Compute is free of charge. However, other related products such as pay-as-you-go Cloud Virtual Machine (CVM) fees may apply.