Loading...
Loading...
Found 440 Skills
Analyzes the AMD GPU memory hierarchy with rocprof-compute, covering the Memory Chart, vL1D and L2 caches, LDS bank conflicts, Infinity Fabric traffic, and the experimental gfx950 memory bandwidth analysis. Use when the user asks about memory bandwidth, cache hit rates, coalescing, LDS conflicts, data locality, HBM traffic, or says a kernel is memory-bound. Not for CUDA tools, Nsight, Windows, host-memory leaks, GPU out-of-memory allocation failures, or system-wide tracing.
Attributes AMD GPU kernel performance counters back to the PyTorch or Triton operators that launched them, using rocprof-compute ML API tracing. Use when the user asks which PyTorch layer or operator is slow, wants to map GPU kernels to model code, profiles a training or inference script, asks about torch.compile and Inductor kernels, or wants to select user-defined ROCTx regions from a traced Python profile. Not for CUDA tools, Windows, compiled non-Python binaries, or starting and stopping a profile at ROCTx ranges.
Reads the rocprof-compute System Speed-of-Light panel to show how close an AMD GPU kernel runs to hardware peak for compute throughput, memory bandwidth, and utilization. Use when the user asks how close to peak a kernel is, what percent of peak it reaches, whether a measured percentage is good enough or normal, whether a utilization number should be higher, or asks for an overall efficiency summary. Not for CUDA tools, Windows, per-instruction analysis, or a datasheet or theoretical-peak question when there is no profile to read.
dstack is an open-source control plane for GPU provisioning and orchestration across GPU clouds, Kubernetes, and on-prem clusters.
Best practices for Docker-based ROS2 development including multi-stage Dockerfiles, docker-compose for multi-container robotic systems, DDS discovery across containers, GPU passthrough for perception, and dev-vs-deploy container patterns. Use this skill when containerizing ROS2 workspaces, setting up docker-compose for robot software stacks, debugging DDS communication between containers, configuring NVIDIA Container Toolkit for GPU workloads, forwarding X11/Wayland for rviz2 and GUI tools, or managing USB device passthrough for cameras and serial devices. Trigger whenever the user mentions Docker with ROS2, docker-compose for robots, Dockerfile for colcon workspaces, container networking for DDS, GPU containers for perception, devcontainer for ROS2, multi-stage builds for ROS2, or deploying ROS2 in containers. Also trigger for CI/CD with Docker-based ROS2 builds, CycloneDDS or FastDDS configuration in containers, shared memory in Docker, or X11 forwarding for rviz2. Covers Humble, Iron, Jazzy, and Rolling distributions across Ubuntu 22.04 and 24.04 base images.
Official Reference Guide for the PPIO Platform, covering LLM API (OpenAI-compatible), Agent Sandbox, GPU (Instances and Serverless), integration, authentication, pricing, rate limiting, and troubleshooting. Suitable for common questions such as 'How to integrate PPIO in specific application scenarios?' and PPIO request failures.
Optimize existing Triton kernels for NVIDIA TileIR backend on Blackwell GPUs (sm_100+). Adds TileIR-specific autotune configs: occupancy, num_ctas, TMA descriptors. Covers kernel classification (dot-related, norm-like, elementwise, reduction), type-specific transformations, and PTX-vs-TileIR benchmarking. Triggered by: "optimize for TileIR", "add TileIR configs", "Blackwell optimization", "TMA descriptors", "2CTA mode", "occupancy tuning". Kernels use standard `import triton`; TileIR activates via ENABLE_TILE=1 when nvtriton is installed.
Converts cuTile Python GPU kernels (@ct.kernel) to cuTile.jl Julia equivalents. Handles kernel syntax translation, 0-indexed to 1-indexed conversion, broadcasting differences, memory layout (row-major to column-major), type system mapping, and launch API differences. Use when converting, porting, or translating cuTile Python kernels to Julia cuTile.jl, or debugging/optimizing existing Julia cuTile translations.
Use Google Colab CLI from a terminal agent to provision CPU, GPU, or TPU runtimes, run Python scripts and notebooks remotely, install packages, move files, export logs, diagnose authentication or allocation failures, and clean up sessions. Use for Colab runtime work, remote GPU/TPU execution, QLoRA or fine-tuning on Colab, notebook execution, Colab CLI troubleshooting, or `colab` command guidance.
Use when making or fixing particle effects and game feel in Godot 4.7: GPUParticles3D or GPUParticles2D explosion, fireball projectile with trail, impact, sparks, smoke, shockwave, ribbon trails, sub-emitters, particle collision, color curve does nothing, particles pop out or vanish, effect too heavy on mobile, CPUParticles fallback for Compatibility or web, overdraw, hit-stop, freeze frame, screen shake, juice.
CuTe Python DSL API reference and implementation patterns for NVIDIA GPU kernel programming. Provides execution model, core API table, key constraints, common patterns, and documentation index. Use when: (1) writing or modifying CuTe DSL kernel code, (2) looking up CuTe DSL API syntax, (3) implementing attention/GEMM/MLA patterns in CuTe DSL, (4) understanding CuTe DSL execution model and compilation pipeline, (5) checking what CuTe DSL can and cannot do.
Run GPU workloads on Modal — training, fine-tuning, inference, batch processing. Zero-config serverless: no SSH, no Docker, auto scale-to-zero. Use when user says "modal run", "modal training", "modal inference", "deploy to modal", "need a GPU", "run on modal", "serverless GPU", or needs remote GPU compute.