skills.rest
xchang1121
xchang1121COMMUNITY
Explore all skills by this author →
GitHub

tilelang-cuda-basics

Teach developers to write TileLang CUDA kernels with shared and global memory.

6|1|Updated Apr 19, 2026
Search Tags:#patterns#gpu#kernel#cuda#memory-management#tilelang#jitted
One-click install
npx skills add https://github.com/xchang1121/AutoResearch-CC-hook --skill tilelang-cuda-basics
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: tilelang-cuda-basics
Source: https://github.com/xchang1121/AutoResearch-CC-hook/tree/main/skills/tilelang-cuda/guides/tilelang-cuda-basics
Command: npx skills add https://github.com/xchang1121/AutoResearch-CC-hook --skill tilelang-cuda-basics

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

TileLang CUDA provides a streamlined pathway for developers to create efficient GPU kernels using a Python-like DSL, reducing complexity in kernel logic and memory management.

Core Features & Use Cases

  • Understand TileLang CUDA core concepts, kernel structure, and memory hierarchy to write optimized kernels.
  • Apply patterns such as elementwise operations and GEMM, enabling practical GPU computation tasks like formatting data, transforming arrays, or performing matrix operations.
  • Use TileLang CUDA to prototype and optimize small to moderate kernels for research and production workflows.

Quick Start

Install TileLang CUDA and run the provided basic example to compile and execute a simple kernel.

Frequently Asked Questions about tilelang-cuda-basics

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

FAQPage Schema
How do I write GPU kernels using TileLang CUDA?▼

TileLang CUDA teaches developers to write GPU kernels by applying syntax rules, kernel structure patterns, and memory hierarchy management using T.Kernel contexts and prim_func abstractions for optimized computation tasks.

What is TileLang CUDA used for in GPU kernel development?▼

TileLang CUDA is used for GPU kernel development across education, research, and production to write and optimize simple to moderately complex kernels, including elementwise operations and GEMM for matrix transformations.

How does memory management work in TileLang CUDA kernels?▼

Memory management in TileLang CUDA kernels involves handling shared and global memory hierarchies, allowing developers to optimize data formatting and array transformations while reducing complexity in kernel logic.

Can I use TileLang CUDA for production GPU computation tasks?▼

Yes, you can use TileLang CUDA to prototype and optimize small to moderate kernels for production GPU computation workflows, applying patterns like elementwise operations and GEMM to perform practical matrix operations.

Do I need prior CUDA experience to use TileLang for GPU kernels?▼

Prior CUDA experience helps, but TileLang CUDA provides a streamlined pathway with a Python-like DSL that reduces complexity in kernel logic, making it accessible for developers prototyping simple to moderately complex kernels.

Related Skills

Explore More Software Engineering Skills→
mindspore-aimindspore-ai
258

tilelang-cuda-patterns

Generate TileLang CUDA kernels from elementwise, reduction, matmul, and GEMV templates.

Official
Intermediate
leonling-llleonling-ll

kernel-perf-analysis

Collect timing, hardware counters, ATT traces, and memory-hierarchy data from AMD GPU kernels.

Community
Advanced
mindspore-aimindspore-ai
258

cuda-c-examples-torch

Compile and call CUDA C kernels from PyTorch using load_inline.

Official
Advanced
pkuppenspkuppens

mojo-gpu-fundamentals

Write GPU kernels in Mojo without CUDA syntax for NVIDIA, AMD, and Apple silicon.

Community
Advanced