Provides low-level GPU device management, memory, stream, and kernel-launching APIs.
npx skills add https://github.com/dongg622/china-ai-chip-skill --skill mxmaca-api
Please help me install this Agent Skill. Skill: mxmaca-api Source: https://github.com/dongg622/china-ai-chip-skill/tree/main/MetaX/mxmaca-api Command: npx skills add https://github.com/dongg622/china-ai-chip-skill --skill mxmaca-api
This Skill provides comprehensive GPU device management, memory handling, stream control, and kernel launching through low-level APIs, simplifying hardware programming tasks.
Use the API functions to select a device, allocate memory, create streams, and launch kernels directly within your development environment.
High-intent search queries and answers about installing and using this skill.
GPU device memory and streams are managed through low-level APIs that handle memory allocation, copying, and stream synchronization to enable efficient hardware acceleration workflows. This Skill provides interfaces for these exact operations.
Custom GPU kernels can be launched using APIs that accept configurable grid, block, shared memory, and stream parameters. This allows tailored computation and flexible deployment within development environments.
Low-level APIs for GPU resource monitoring and performance profiling require device query interfaces and property access functions. This Skill provides device management capabilities to query active GPUs and access device properties for optimized resource allocation.
Execution streams and events are synchronized using dedicated control APIs for fine-grained GPU scheduling. This Skill enables creating and synchronizing streams and events to ensure proper execution order and dependency management.
Allocating and freeing both device and host memory is supported through dedicated memory operation APIs. This Skill ensures efficient memory handling for high-performance computing by providing allocation, copy, and free operations.
jayll1303Build GPU-enabled Docker containers with NGC base images and uv.
mtsatryanProfile CPU, memory, I/O, and latency to identify performance bottlenecks.
hung-phanCreate high-performance GPU kernels for deep learning using Python.
jstzwjGenerates GPU kernels for CUDA, HIP, Metal, and CPU via Python.
