Run GPU-accelerated batch cosine similarity and matrix multiplication with WGPU.
npx skills add https://github.com/GuitarAlchemist/ix --skill ix-gpu
Please help me install this Agent Skill. Skill: ix-gpu Source: https://github.com/GuitarAlchemist/ix/tree/main/.claude/skills/ix-gpu Command: npx skills add https://github.com/GuitarAlchemist/ix --skill ix-gpu
The ix-gpu skill addresses the need for GPU-accelerated operations on large datasets, including batch cosine similarity search, matrix multiplication, and more.
Run GPU-accelerated cosine similarity search with the ix-gpu skill by executing the following command: use ix-gpu::similarity::GpuCosineSimilarity;
High-intent search queries and answers about installing and using this skill.
Batch cosine similarity search can be accelerated using GPU computation via WGPU, which speeds up processing large vector collections. It includes CPU fallbacks for scenarios where a GPU is not available.
GPU acceleration for matrix multiplication is supported to handle large-scale data operations efficiently. Using WGPU, it processes matrices with hundreds of rows and columns significantly faster than CPU-only methods.
Large-scale vector search can still run without a dedicated GPU by utilizing the built-in CPU fallbacks. This ensures batch similarity searches remain operational even when WGPU hardware acceleration is not accessible.
WGPU supports batch operations for quaternion and sedenion transformations alongside matrix multiplication and similarity searches. This allows efficient GPU computation across large batches of complex numerical data.
You can start GPU computation for batch vector search by importing the GPU cosine similarity module using the command `use ix-gpu::similarity::GpuCosineSimilarity;` to accelerate your vector collections.
You should avoid GPU-accelerated computation for very small-scale numerical operations where the overhead of WGPU initialization outweighs the speedup benefits, and instead rely on the included CPU fallbacks.
cloudlinessEnable fast, scalable vector similarity search with FAISS indices on CPU or GPU.
vicharanashalaPractice 2x2 matrix operations with adaptive difficulty and real-time validation.
jarmak-personalWire new spatial operations into vibeSpatial's dispatch system with type coercion.
robertbr123Implement vector similarity search with FAISS and NumPy.
