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.
Duration 14 hours
Course Outline
Introductory Overview of Huawei's AI Ecosystem
- An introduction to Ascend AI hardware, including the 310, 910, and 910B series
- Key high-level components: MindSpore, CANN, and AscendCL
- Strategic industry positioning and core architectural principles
The Function of CANN within Huawei's AI Framework
- Defining CANN: SDK objectives and internal structural layers
- ATC, TBE, and AscendCL: Processes for compiling and executing models
- How CANN enhances inference optimization and deployment capabilities
MindSpore: Architecture and Overview
- Workflows for training and inference within the MindSpore environment
- Concepts of Graph mode, PyNative, and hardware abstraction
- Seamless integration with Ascend NPUs through the CANN backend
Navigating the AI Lifecycle on Ascend: From Training to Deployment
- Creating models in MindSpore or converting from external frameworks
- Exporting and compiling models utilizing ATC tools
- Deploying to Ascend hardware using OM models and AscendCL interfaces
Benchmarking Against Competing AI Stacks
- MindSpore versus PyTorch and TensorFlow: Distinct focuses and market positioning
- Comparative analysis of deployment workflows on Ascend versus GPU-based ecosystems
- Evaluating opportunities and constraints for enterprise adoption
Real-World Enterprise Integration Scenarios
- Application cases in smart manufacturing, government AI initiatives, and telecommunications
- Key considerations regarding scalability, regulatory compliance, and ecosystem support
- Implementing cloud/on-premises hybrid deployments with the Huawei stack
Concluding Summary and Recommended Next Steps
Requirements
- General familiarity with AI workflows or platform architectural concepts
- Foundational knowledge of model training and deployment processes
- No prior hands-on experience with CANN or MindSpore is necessary
Target Audience
- AI platform evaluators and infrastructure architects
- AI/ML DevOps engineers and pipeline integration specialists
- Technology leaders and strategic decision-makers
