Ascend Cloud Service

Vast Computing Power

Instant access to immense computing power; 10,000-card clusters.

Efficient Long-Term Training

30+ days of uninterrupted training on 1,000+ cards; training tasks auto-recovered in less than 30 minutes.

Complete Toolchains

E2E, cloud-based toolchains; configuration-free and available out-of-the-box; self-service migration for mainstream scenarios.

Full-Stack Ecosystem

Adapted to support major open-source models; 100,000+ assets available in AI Gallery.

Ascend Cloud Service Architecture

Ascend Cloud Service Architecture

Terms & Conditions


活动时间: 2020年8月12日-2020年9月11日

Why Ascend Cloud Service?

Industry-Leading Ascend Architecture

Industry-Leading Ascend Architecture

  • Using Da Vinci architecture, Ascend offers 30% more cost-effective AI computing power than competitors.

  • MindSpore accelerates the tuning of a 100-billion parameter model by 60%.

Efficient Migration Toolchain

Efficient Migration Toolchain

  • For typical scenarios, the E2E migration toolchain can migrate an AI application to a production environment in less than two weeks.

  • Easy-to-use training and inference migration tools enable self-service migration.

Improved Resource Utilization

Improved Resource Utilization

  • Unified resource scheduling and optimized allocation policies achieve 90% overall resource utilization.

  • Elastic scheduling, along with converged resource scheduling for training and inference, ensures resource provisioning in less than 30 minutes.

A Robust AI Ecosystem

A Robust AI Ecosystem

  • AI Gallery aggregates more than 100,000 domain-specific AI assets.

  • Major open-source models are supported and are specially adapted to work smoothly with Ascend AI.

Use Cases

Foundation Model
Full Support for Third-Party Open-Source Foundation Models, Fast Service Rollout for Customers
E2E Toolchain for Out-Of-The-Box Foundation Models

This toolchain covers cloud-based data cleansing, model fine-tuning, deployment, prompt engineering, evaluation, and agents. It accelerates application development from foundation models.

Ascend-native, cloud-based AI agents, such as search agents and big data agents, allow users to integrate the capabilities of AI foundation models into existing cloud services or components with zero migration.

Mainstream AI development frameworks are supported, ensuring seamless migration of applications powered by non-Ascend architectures.

Ascend-compatible Open-Source Foundation Models

Major open-source foundation models have been specially adapted to Ascend architecture, delivering higher accuracy and performance than competitors.

Inner-source code, container images, and performance parameters are fully accessible, allowing for informed technology selection. Migrating an open-source foundation model to Ascend has been accelerated from months to just days.

Ascend Toolchains for All Major Open-Source Foundation Models

Both the Ascend toolchain and native toolchain are available for open-source foundation models, facilitating model tuning and toolchain integration. This allows you to choose the tools that best suit your needs.

Ascend Chips Significantly Improve AIGC Model Performance
Model Conversion

Models can be converted to a MindSpore-compatible format in just a few minutes.

Auto Tuning

Graph optimization is fully automated. During this process, model tuning policies are automatically generated, verified on NPUs based on feedback, and continuously optimized in an iterative manner.

Performance and Accuracy Verification

Tools are used to quantitatively analyze the performance and accuracy of models after they are migrated to Ascend, ensuring zero losses in performance or accuracy.

Autonomous Driving
Efficient Training of Autonomous Driving Models on PB-Scale Data, Faster Innovation and Iteration
In-Depth Optimization of Sensing and Simulation Algorithms

Algorithms along the entire chain of autonomous driving, particularly those related to sensing, route planning, control, simulation, and generation, have been optimized extensively, leading to a substantial enhancement in performance.

Large-Scale Distributed Training

Supports distributed training on petabytes of data.

Compute Upgrade

Instant access to immense Ascend AI computing power accelerates algorithm development and iteration.

Content Moderation
Ascend AI Cloud Service for Content Moderation

Ascend AI Cloud Service offers robust, proven solutions for content moderation and fast migration (of CV workloads) to Ascend, accommodating the compute and business continuity needs of content platforms and service providers.

Fast Migration Evaluation

The Ascend migration toolchain can quickly analyze model operators, diagnose model accuracy, test and optimize model performance, and convert model formats.


Ascend chips boost the performance of content moderation models, providing higher performance than that of competitors.

AI Gallery