Performance Optimization on Ascend, Biren, and Cambricon Training Course
Ascend, Biren, and Cambricon are prominent AI hardware platforms in China, each providing distinct acceleration and profiling tools for production-scale AI workloads.
This instructor-led, live training (available online or onsite) is designed for advanced-level AI infrastructure and performance engineers who seek to optimise model inference and training workflows across multiple Chinese AI chip platforms.
By the conclusion of this training, participants will be able to:
- Benchmark models on Ascend, Biren, and Cambricon platforms.
- Identify system bottlenecks and memory/compute inefficiencies.
- Apply graph-level, kernel-level, and operator-level optimisations.
- Tune deployment pipelines to enhance throughput and reduce latency.
Course Format
- Interactive lectures and discussions.
- Practical application of profiling and optimisation tools on each platform.
- Guided exercises focused on real-world tuning scenarios.
Course Customisation Options
- To request a bespoke training session tailored to your performance environment or model type, please contact us to arrange.
Course Outline
Performance Concepts and Metrics
- Latency, throughput, power usage, resource utilization
- System vs model-level bottlenecks
- Profiling for inference vs training
Profiling on Huawei Ascend
- Using CANN Profiler and MindInsight
- Kernel and operator diagnostics
- Offload patterns and memory mapping
Profiling on Biren GPU
- Biren SDK performance monitoring features
- Kernel fusion, memory alignment, and execution queues
- Power and temperature-aware profiling
Profiling on Cambricon MLU
- BANGPy and Neuware performance tools
- Kernel-level visibility and log interpretation
- MLU profiler integration with deployment frameworks
Graph and Model-Level Optimisation
- Graph pruning and quantization strategies
- Operator fusion and computational graph restructuring
- Input size standardization and batch tuning
Memory and Kernel Optimisation
- Optimizing memory layout and reuse
- Efficient buffer management across chipsets
- Kernel-level tuning techniques per platform
Cross-Platform Best Practices
- Performance portability: abstraction strategies
- Building shared tuning pipelines for multi-chip environments
- Example: tuning an object detection model across Ascend, Biren, and MLU
Summary and Next Steps
Requirements
- Experience working with AI model training or deployment pipelines
- Understanding of GPU/MLU compute principles and model optimisation
- Basic familiarity with performance profiling tools and metrics
Audience
- Performance engineers
- Machine learning infrastructure teams
- AI system architects
Open Training Courses require 5+ participants.
Performance Optimization on Ascend, Biren, and Cambricon Training Course - Booking
Performance Optimization on Ascend, Biren, and Cambricon Training Course - Enquiry
Performance Optimization on Ascend, Biren, and Cambricon - Consultancy Enquiry
Provisional Upcoming Courses (Require 5+ participants)
Related Courses
Developing AI Applications with Huawei Ascend and CANN
21 HoursHuawei Ascend is a family of AI processors designed for high-performance inference and training.
This instructor-led, live training (online or onsite) is aimed at intermediate-level AI engineers and data scientists who wish to develop and optimize neural network models using Huawei’s Ascend platform and the CANN toolkit.
By the end of this training, participants will be able to:
- Set up and configure the CANN development environment.
- Develop AI applications using MindSpore and CloudMatrix workflows.
- Optimise performance on Ascend NPUs using custom operators and tiling.
- Deploy models to edge or cloud environments.
Format of the Course
- Interactive lecture and discussion.
- Hands-on use of Huawei Ascend and CANN toolkit in sample applications.
- Guided exercises focused on model building, training, and deployment.
Course Customisation Options
- To request a customised training for this course based on your infrastructure or datasets, please contact us to arrange.
Deploying AI Models with CANN and Ascend AI Processors
14 HoursCANN (Compute Architecture for Neural Networks) serves as Huawei’s AI compute stack, designed for the deployment and optimisation of AI models on Ascend AI processors.
This instructor-led, live training (delivered online or onsite) is tailored for intermediate-level AI developers and engineers who want to efficiently deploy trained AI models onto Huawei Ascend hardware using the CANN toolkit and tools such as MindSpore, TensorFlow, or PyTorch.
Upon completing this training, participants will be able to:
- Grasp the CANN architecture and its significance within the AI deployment pipeline.
- Convert and adapt models from popular frameworks into Ascend-compatible formats.
- Utilise tools such as ATC, OM model conversion, and MindSpore for edge and cloud inference.
- Diagnose deployment challenges and optimise performance on Ascend hardware.
Course Format
- Interactive lectures and demonstrations.
- Hands-on lab exercises using CANN tools alongside Ascend simulators or devices.
- Practical deployment scenarios grounded in real-world AI models.
Course Customisation Options
- To request customised training for this course, please get in touch with us to arrange.
AI Inference and Deployment with CloudMatrix
21 HoursCloudMatrix is Huawei’s unified AI development and deployment platform designed to support scalable, production-grade inference pipelines.
This instructor-led, live training (online or onsite) is aimed at beginner-level to intermediate-level AI professionals who wish to deploy and monitor AI models using the CloudMatrix platform with CANN and MindSpore integration.
By the end of this training, participants will be able to:
- Use CloudMatrix for model packaging, deployment, and serving.
- Convert and optimize models for Ascend chipsets.
- Set up pipelines for real-time and batch inference tasks.
- Monitor deployments and tune performance in production settings.
Format of the Course
- Interactive lecture and discussion.
- Hands-on use of CloudMatrix with real deployment scenarios.
- Guided exercises focused on conversion, optimization, and scaling.
Course Customization Options
- To request a customized training for this course based on your AI infrastructure or cloud environment, please contact us to arrange.
GPU Programming on Biren AI Accelerators
21 HoursBiren AI Accelerators are high-performance GPUs designed for AI and HPC workloads with support for large-scale training and inference.
This instructor-led, live training (online or onsite) is aimed at intermediate-level to advanced-level developers who wish to program and optimize applications using Biren’s proprietary GPU stack, with practical comparisons to CUDA-based environments.
By the end of this training, participants will be able to:
- Understand Biren GPU architecture and memory hierarchy.
- Set up the development environment and use Biren’s programming model.
- Translate and optimize CUDA-style code for Biren platforms.
- Apply performance tuning and debugging techniques.
Format of the Course
- Interactive lecture and discussion.
- Hands-on use of Biren SDK in sample GPU workloads.
- Guided exercises focused on porting and performance tuning.
Course Customization Options
- To request a customized training for this course based on your application stack or integration needs, please contact us to arrange.
Cambricon MLU Development with BANGPy and Neuware
21 HoursCambricon MLUs (Machine Learning Units) are dedicated AI chips engineered for high-performance inference and training in both edge and data centre environments.
This instructor-led live training, available online or onsite, is designed for intermediate-level developers looking to build and deploy AI models on Cambricon MLU hardware using the BANGPy framework and Neuware SDK.
Upon completion of this training, participants will be able to:
- Set up and configure the BANGPy and Neuware development environments.
- Develop and optimise Python- and C++-based models specifically for Cambricon MLUs.
- Deploy models to edge and data centre devices running the Neuware runtime.
- Integrate machine learning workflows with MLU-specific acceleration features.
Course Format
- Interactive lectures and discussions.
- Hands-on development and deployment using BANGPy and Neuware.
- Guided exercises focusing on optimisation, integration, and testing.
Customisation Options
- To request tailored training for this course based on your specific Cambricon device model or use case, please contact us to arrange.
Introduction to CANN for AI Framework Developers
7 HoursCANN (Compute Architecture for Neural Networks) serves as Huawei’s AI computing toolkit, designed to compile, optimise, and deploy AI models across Ascend AI processors.
This instructor-led live training, available both online and onsite, targets beginner-level AI developers keen to grasp how CANN integrates into the model lifecycle—from training to deployment—and how it interfaces with frameworks such as MindSpore, TensorFlow, and PyTorch.
Upon completion of this training, participants will be able to:
- Grasp the purpose and architecture of the CANN toolkit.
- Configure a development environment utilising CANN and MindSpore.
- Convert and deploy a simple AI model onto Ascend hardware.
- Build foundational knowledge for future CANN optimisation or integration initiatives.
Course Format
- Interactive lectures and discussions.
- Practical labs focused on simple model deployment.
- Step-by-step walkthroughs of the CANN toolchain and integration points.
Customisation Options
- To arrange a tailored training session for this course, please get in touch with us.
CANN for Edge AI Deployment
14 HoursHuawei's Ascend CANN toolkit empowers robust AI inference on edge devices like the Ascend 310. It supplies crucial tools for compiling, optimising, and deploying models in environments where compute power and memory are limited.
This instructor-led live training (available online or onsite) is designed for intermediate-level AI developers and integrators keen to deploy and optimise models on Ascend edge devices using the CANN toolchain.
By the end of this training, participants will be able to:
- Prepare and convert AI models for the Ascend 310 using CANN tools.
- Construct lightweight inference pipelines using MindSpore Lite and AscendCL.
- Optimise model performance for constrained compute and memory environments.
- Deploy and monitor AI applications in real-world edge scenarios.
Course Format
- Interactive lectures and demonstrations.
- Hands-on lab exercises featuring edge-specific models and scenarios.
- Live deployment examples on virtual or physical edge hardware.
Course Customisation Options
- To request a customised training course, please contact us to arrange.
Understanding Huawei’s AI Compute Stack: From CANN to MindSpore
14 HoursHuawei’s AI stack — spanning from the low-level CANN SDK to the high-level MindSpore framework — provides a tightly integrated environment for AI development and deployment, specifically optimised for Ascend hardware.
This instructor-led, live training session (available online or onsite) is designed for beginner to intermediate-level technical professionals who wish to understand how the CANN and MindSpore components collaborate to support AI lifecycle management and infrastructure decision-making.
By the conclusion of this training, participants will be able to:
- Comprehend the layered architecture of Huawei’s AI compute stack.
- Identify how CANN facilitates model optimisation and hardware-level deployment.
- Evaluate the MindSpore framework and its toolchain in comparison to industry alternatives.
- Position Huawei’s AI stack within enterprise or cloud/on-premises environments.
Course Format
- Interactive lectures and discussions.
- Live system demonstrations and case-based walkthroughs.
- Optional guided labs covering the model flow from MindSpore to CANN.
Course Customisation Options
- To request a customised training programme for this course, please contact us to make arrangements.
Optimizing Neural Network Performance with CANN SDK
14 HoursThe CANN SDK (Compute Architecture for Neural Networks) serves as Huawei’s foundational AI compute platform, empowering developers to fine-tune and optimise the performance of neural networks deployed on Ascend AI processors.
This instructor-led live training (available online or onsite) is designed for advanced AI developers and system engineers looking to enhance inference performance using CANN’s advanced toolset, including the Graph Engine, TIK, and custom operator development capabilities.
Upon completion of this training, participants will be able to:
- Gain a clear understanding of CANN’s runtime architecture and performance lifecycle.
- Utilise profiling tools and the Graph Engine for performance analysis and optimisation.
- Create and optimise custom operators using TIK and TVM.
- Address memory bottlenecks and enhance model throughput.
Format of the Course
- Interactive lectures and discussions.
- Hands-on labs featuring real-time profiling and operator tuning.
- Optimisation exercises using edge-case deployment examples.
Course Customization Options
- To request a tailored training session for this course, please contact us to make arrangements.
CANN SDK for Computer Vision and NLP Pipelines
14 HoursThe CANN SDK (Compute Architecture for Neural Networks) equips developers with powerful deployment and optimization tools for real-time AI applications in computer vision and NLP, particularly on Huawei Ascend hardware.
This instructor-led, live training (available online or onsite) is designed for intermediate-level AI practitioners looking to build, deploy, and optimize vision and language models using the CANN SDK for production-grade use cases.
Upon completion of this training, participants will be able to:
- Deploy and optimize CV and NLP models leveraging CANN and AscendCL.
- Utilise CANN tools to convert models and seamlessly integrate them into live pipelines.
- Optimise inference performance for tasks such as detection, classification, and sentiment analysis.
- Construct real-time CV/NLP pipelines tailored for edge or cloud-based deployment scenarios.
Format of the Course
- Interactive lecture and demonstration.
- Hands-on lab sessions focused on model deployment and performance profiling.
- Live pipeline design incorporating real-world CV and NLP use cases.
Course Customization Options
- To request a customized training session for this course, please contact us to arrange.
Building Custom AI Operators with CANN TIK and TVM
14 HoursLeveraging CANN TIK (Tensor Instruction Kernel) alongside Apache TVM allows for sophisticated optimisation and customisation of AI model operators tailored to Huawei Ascend hardware.
This instructor-led, live training session (available online or onsite) is designed for advanced system developers looking to create, deploy, and fine-tune custom operators for AI models using CANN's TIK programming model and TVM compiler integration.
Upon completion of this training, participants will be equipped to:
- Write and test custom AI operators utilizing the TIK DSL for Ascend processors.
- Integrate bespoke operators into the CANN runtime and execution graph.
- Employ TVM for operator scheduling, auto-tuning, and benchmarking.
- Debug and optimise instruction-level performance for bespoke computation patterns.
Course Format
- Interactive lectures and live demonstrations.
- Practical coding exercises for operators using TIK and TVM pipelines.
- Testing and tuning on Ascend hardware or simulators.
Customisation Options
- To request a tailored training version of this course, please contact us to make arrangements.
Migrating CUDA Applications to Chinese GPU Architectures
21 HoursIndigenous GPU architectures in China, including Huawei Ascend, Biren, and Cambricon MLUs, provide viable alternatives to CUDA, specifically designed for the domestic artificial intelligence and high-performance computing (HPC) sectors.
This instructor-led, live training session (available online or onsite) targets advanced GPU developers and infrastructure specialists seeking to migrate and optimise existing CUDA applications for deployment on Chinese hardware platforms.
Upon completion of this training, participants will be capable of:
- Evaluating the compatibility of current CUDA workloads with domestic chip alternatives.
- Translating CUDA codebases into Huawei CANN, Biren SDK, and Cambricon BANGPy environments.
- Comparing performance metrics and identifying key areas for optimisation across different platforms.
- Navigating practical challenges associated with cross-architecture support and deployment.
Course Delivery Format
- Interactive lectures and discussions.
- Practical labs involving code translation and performance benchmarking.
- Guided exercises focusing on multi-GPU adaptation strategies.
Customisation Options
- To arrange customised training for this course tailored to your specific platform or CUDA project, please contact us.
