Get in Touch

Course Outline

Performance Concepts and Metrics

  • Latency, throughput, power consumption, and resource utilisation
  • Distinguishing between system-level and model-level bottlenecks
  • Profiling strategies for inference versus training

Profiling on Huawei Ascend

  • Leveraging CANN Profiler and MindInsight
  • Diagnosing kernels and operators
  • Managing offload patterns and memory mapping

Profiling on Biren GPU

  • Utilising Biren SDK performance monitoring capabilities
  • Addressing kernel fusion, memory alignment, and execution queues
  • Conducting power and temperature-aware profiling

Profiling on Cambricon MLU

  • Using BANGPy and Neuware performance tools
  • Gaining kernel-level visibility and interpreting logs
  • Integrating the MLU profiler with deployment frameworks

Graph and Model-Level Optimisation

  • Strategies for graph pruning and quantization
  • Operator fusion and restructuring computational graphs
  • Standardising input sizes and tuning batch parameters

Memory and Kernel Optimisation

  • Refining memory layout and reuse strategies
  • Efficient buffer management across different chipsets
  • Applying platform-specific kernel-level tuning techniques

Cross-Platform Best Practices

  • Achieving performance portability through abstraction strategies
  • Developing shared tuning pipelines for multi-chip environments
  • Case study: tuning an object detection model across Ascend, Biren, and MLU

Summary and Next Steps

Requirements

  • Practical experience with AI model training or deployment pipelines
  • Knowledge of GPU/MLU compute principles and model optimisation
  • Foundational understanding of performance profiling tools and key metrics

Target Audience

  • Performance engineers
  • Machine learning infrastructure teams
  • AI system architects
 21 Hours

Number of participants


Price per participant

Provisional Upcoming Courses (Require 5+ participants)

Related Categories