Get in Touch

Course Outline

Introduction to the Huawei Ascend Platform

  • Overview of the Ascend ecosystem and architecture.
  • Introduction to CANN and MindSpore.
  • Industry relevance and use cases.

Establishing the Development Environment

  • Installation of MindSpore and the CANN toolkit.
  • Using CloudMatrix and ModelArts for project orchestration.
  • Verifying the environment with sample models.

Model Development with MindSpore

  • Defining and training models within MindSpore.
  • Dataset formatting and data pipelines.
  • Exporting models to formats compatible with Ascend.

Performance Optimization on Ascend

  • Custom kernels and operator fusion.
  • AI Core scheduling and tiling strategies.
  • Profiling and benchmarking tools.

Deployment Strategies

  • Tradeoffs between edge and cloud deployment.
  • Using the MindX SDK for deployment.
  • Integration with CloudMatrix workflows.

Debugging and Monitoring

  • Tracing with AiD and Profiler.
  • Resolving runtime failures.
  • Monitoring throughput and resource usage.

Lab Integration and Case Study

  • Complete pipeline development using MindSpore.
  • Lab: Develop, optimize, and deploy a model on Ascend.
  • Comparing performance with other platforms.

Summary and Next Steps

Requirements

  • Knowledge of neural networks and AI workflows.
  • Proficiency in Python programming.
  • Familiarity with pipelines for model training and deployment.

Target Audience

  • AI engineers.
  • Data scientists utilizing the Huawei AI stack.
  • ML developers employing Ascend and MindSpore.
 21 Hours

Number of participants


Price per participant

Testimonials (1)

Upcoming Courses

Related Categories