Thank you for sending your enquiry! One of our team members will contact you shortly.
Thank you for sending your booking! One of our team members will contact you shortly.
Course Outline
Introduction to Huawei’s AI Ecosystem
- Overview of Ascend AI hardware: models 310, 910, and 910B.
- Key components: MindSpore, CANN, and AscendCL.
- Industry positioning and core architectural principles.
The Role of CANN in Huawei’s AI Stack
- Understanding CANN: SDK purpose and internal layers.
- Roles of ATC, TBE, and AscendCL in compiling and executing models.
- How CANN enables inference optimization and deployment.
Overview and Architecture of MindSpore
- Training and inference workflows within MindSpore.
- Graph mode, PyNative, and hardware abstraction techniques.
- Integration with Ascend NPUs via the CANN backend.
AI Lifecycle on Ascend: From Training to Deployment
- Creating models in MindSpore or converting them from other frameworks.
- Exporting and compiling models using ATC.
- Deploying on Ascend hardware using OM models and AscendCL.
Comparison with Alternative AI Stacks
- MindSpore versus PyTorch and TensorFlow: focus areas and positioning.
- Deployment workflows on Ascend compared to GPU-based stacks.
- Opportunities and limitations for enterprise adoption.
Enterprise Integration Scenarios
- Use cases in smart manufacturing, government AI, and telecom sectors.
- Considerations regarding scalability, compliance, and ecosystem.
- Hybrid cloud/on-premises deployment strategies using the Huawei stack.
Summary and Next Steps
Requirements
- Familiarity with AI workflows or platform architectures.
- Basic knowledge of model training and deployment processes.
- No prior hands-on experience with CANN or MindSpore is required.
Target Audience
- AI platform evaluators and infrastructure architects.
- AI/ML DevOps professionals and pipeline integrators.
- Technology managers and decision-makers.
14 Hours
