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 21 hours
Course Outline
Introduction to the Chinese AI GPU Landscape
- Comparative analysis of Huawei Ascend, Biren, and Cambricon MLU
- Contrasting CUDA with CANN, Biren SDK, and BANGPy models
- Market trends and vendor ecosystem dynamics
Migration Preparation
- Evaluating the current CUDA codebase
- Defining target platforms and specific SDK versions
- Setting up the toolchain and development environment
Code Conversion Strategies
- Translating CUDA memory access patterns and kernel logic
- Aligning compute grid and thread models
- Choosing between automated and manual translation methods
Platform-Specific Development
- Leveraging Huawei CANN operators and custom kernels
- Utilizing the Biren SDK conversion workflow
- Reconstructing models using BANGPy (Cambricon)
Inter-Platform Testing and Tuning
- Profiling execution metrics on each target architecture
- Comparing memory tuning and parallel execution behaviors
- Monitoring performance and iterating on results
Managing Heterogeneous GPU Environments
- Implementing hybrid deployments across multiple architectures
- Designing fallback mechanisms and device detection protocols
- Establishing abstraction layers to ensure code maintainability
Practical Case Studies and Industry Standards
- Migrating vision and NLP models to Ascend or Cambricon
- Adapting inference pipelines for Biren clusters
- Resolving version conflicts and API discrepancies
Conclusion and Future Directions
Requirements
- Practical experience in CUDA or GPU-based application development
- Working knowledge of GPU memory models and compute kernels
- Exposure to AI model deployment or acceleration processes
Target Audience
- GPU Developers
- System Architects
- Software Porting Specialists
