skills.rest
dongg622
dongg622COMMUNITY
Explore all skills by this author →
GitHub

cuda-adapt

Migrate CUDA applications to the海光DCU platform with automated code conversion and API mapping.

Updated Aug 27, 2026
Search Tags:#performance optimization#code migration#cuda#hip#dcU#hipify#kernel adaptation
One-click install
npx skills add https://github.com/dongg622/china-ai-chip-skill --skill cuda-adapt
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: cuda-adapt
Source: https://github.com/dongg622/china-ai-chip-skill/tree/main/Hygon/cuda-adapt
Command: npx skills add https://github.com/dongg622/china-ai-chip-skill --skill cuda-adapt

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill simplifies the complex process of migrating CUDA applications to the海光DCU platform by providing comprehensive adaptation guidance and tools.

Core Features & Use Cases

  • CUDA Code Migration: Supports code conversion, API replacement, and debugging for CUDA programs migrating to DCU.
  • Performance Optimization: Offers best practices for tuning and enhancing CUDA code performance on DCU.
  • Use Case: A developer upgrading a machine learning pipeline from CUDA to DCU can follow this Skill to ensure compatibility and optimize performance with minimal manual effort.

Quick Start

Launch the migration process by reviewing the provided CUDA code and applying the recommended API mappings and performance tips directly using the scripts and instructions in this Skill repository.

Frequently Asked Questions about cuda-adapt

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I migrate CUDA code to the Hygon DCU platform?▼

This Skill streamlines migrating CUDA code to the Hygon DCU platform through automated code conversion, API mapping, and debugging guidance to establish full hardware compatibility with minimal manual intervention.

Can I use hipify for kernel adaptation to DCU?▼

Yes, this Skill supports kernel adaptation by utilizing hipify methodologies alongside API replacement to translate CUDA programs for DCU deployment, ensuring accurate code compatibility and functional execution.

What is the best way to optimize CUDA performance on DCU?▼

Optimizing CUDA performance on DCU involves applying the best practices and tuning guidance consolidated in this Skill repository, which helps developers enhance machine learning pipeline efficiency after migration.

Do I need Python scripts to debug CUDA to DCU code migration?▼

Yes, you need the included Python scripts to execute the debugging and code conversion steps for CUDA to DCU migration, relying on the provided documentation for effective deployment and troubleshooting.

Does this Skill support API replacement for machine learning pipelines moving to DCU?▼

Yes, this Skill supports API replacement and code conversion specifically for upgrading machine learning pipelines from CUDA to DCU, ensuring developers achieve compatibility and optimized performance.

Related Skills

Explore More Software Engineering Skills→
michaeldubumichaeldubu

perf-analyzer

Parse and analyze CUDA, CPU, and memory benchmark outputs across CSV, JSON, and text formats.

Community
Advanced
pluginagentmarketplacepluginagentmarketplace
40

java-performance

Optimize JVM performance with GC tuning, profiling, and JMH benchmarking.

Community
Advanced
PriyanshKuniyalPriyanshKuniyal

python-performance-optimization

Profile Python code with cProfile and memory profilers to identify bottlenecks.

Community
Advanced
tensormuxtensormux
54

Port CUDA Kernel to HIP

Port CUDA kernels to HIP for ROCm-enabled AMD GPUs.

Official
Advanced