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
Overview of CANN Optimization Capabilities for Government
- How Inference Performance is Managed in CANN
- Optimization Goals for Edge and Embedded AI Systems in the Public Sector
- Understanding AI Core Utilization and Memory Allocation for Efficient Resource Management
Using Graph Engine for Analysis for Government
- Introduction to the Graph Engine and Execution Pipeline for Enhanced Operational Efficiency
- Visualizing Operator Graphs and Runtime Metrics to Identify Performance Bottlenecks
- Modifying Computational Graphs for Optimization in Public Sector Applications
Profiling Tools and Performance Metrics for Government
- Using the CANN Profiling Tool (profiler) for Workload Analysis in Federal Systems
- Analyzing Kernel Execution Time and Identifying Bottlenecks to Improve System Performance
- Memory Access Profiling and Tiling Strategies for Efficient Data Management
Custom Operator Development with TIK for Government
- Overview of TIK and the Operator Programming Model for Custom AI Solutions
- Implementing a Custom Operator Using the TIK Domain-Specific Language (DSL) for Government Applications
- Testing and Benchmarking Operator Performance to Ensure Reliability and Efficiency
Advanced Operator Optimization with TVM for Government
- Introduction to TVM Integration with CANN for Enhanced Computational Efficiency
- Auto-Tuning Strategies for Computational Graphs to Optimize Resource Utilization
- When and How to Switch Between TVM and TIK for Optimal Performance in Government Systems
Memory Optimization Techniques for Government
- Managing Memory Layout and Buffer Placement to Enhance System Efficiency
- Techniques to Reduce On-Chip Memory Consumption for Sustainable Operations
- Best Practices for Asynchronous Execution and Reuse to Improve Performance in Public Sector Deployments
Real-World Deployment and Case Studies for Government
- Case Study: Performance Tuning for Smart City Camera Pipeline to Enhance Public Safety
- Case Study: Optimizing Autonomous Vehicle Inference Stack for Efficient Transportation Solutions
- Guidelines for Iterative Profiling and Continuous Improvement in Government AI Projects
Summary and Next Steps for Government
Requirements
- Comprehensive knowledge of deep learning model architectures and training workflows for government applications
- Experience in deploying models using CANN, TensorFlow, or PyTorch in a governmental context
- Familiarity with Linux command-line interface (CLI), shell scripting, and Python programming for government systems
Audience
- AI performance engineers working in the public sector
- Inference optimization specialists supporting government projects
- Developers focused on edge AI or real-time systems for government use
14 Hours
