skills.rest
matlab
matlabOFFICIAL
Explore all skills by this author →
WebsiteGitHub

matlab-setup-gpu

Detect and validate GPU availability for MATLAB GPU computing workflows.

995|122|Updated Apr 3, 2026
Search Tags:#GPU#Troubleshooting#Validation#matlab#gpudevice#parallel computing toolbox#driver diagnostics
One-click install
npx skills add https://github.com/matlab/matlab-agentic-toolkit --skill matlab-setup-gpu
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: matlab-setup-gpu
Source: https://github.com/matlab/matlab-agentic-toolkit/tree/main/skills-catalog/parallel-computing/matlab-setup-gpu
Command: npx skills add https://github.com/matlab/matlab-agentic-toolkit --skill matlab-setup-gpu

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

MATLAB users often cannot tell whether their GPU is usable for GPU computing, or why calls to gpuDevice and gpuArray fail. This Skill provides a structured diagnostic workflow to detect compatible GPUs, validate them, and resolve common failure modes such as missing licenses, outdated drivers, and Windows WDDM driver model issues.

Core Features & Use Cases

  • GPU Availability Detection: Uses canUseGPU, gpuDeviceTable, and gpuDevice to check whether a licensed, usable GPU exists and list all visible devices.
  • Structured Diagnostics: Runs validateGPU to test each layer (toolbox, license, driver, device) and maps failures to specific resolutions via a troubleshooting table.
  • Driver Model Analysis: On Windows, inspects DriverModel and KernelExecutionTimeout to identify WDDM-related kernel timeout risks and suggest remediation.
  • Use Case: A user gets "unable to find a supported GPU device" when running gpuArray. The Skill walks through canUseGPU, validateGPU, and gpuDeviceTable to pinpoint whether the cause is licensing, driver version, or device visibility, then directs the user to the GPU Computing Requirements page for their MATLAB release.

Quick Start

Ask the agent to check whether your system has a compatible GPU for MATLAB and diagnose why gpuDevice is failing.

Frequently Asked Questions about matlab-setup-gpu

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

FAQPage Schema
How do I check if my GPU works with MATLAB?▼

Run canUseGPU in MATLAB, which returns true if Parallel Computing Toolbox is licensed and a usable GPU device exists. Then use gpuDeviceTable to list all visible GPUs and gpuDevice(N) to select and inspect a specific device.

Why does gpuDevice fail with "unable to find a supported GPU device"?▼

This error can stem from an unlicensed Parallel Computing Toolbox, an outdated NVIDIA driver, a prohibited or exclusive compute mode, or the GPU not being visible to MATLAB. Run validateGPU to test each layer and identify which check fails.

What MATLAB version do I need for validateGPU?▼

validateGPU requires MATLAB R2024b or later with Parallel Computing Toolbox installed. Check ver('parallel') to confirm the toolbox is installed and its version is 24.2 or higher.

Why does my GPU kernel get killed mid-computation on Windows?▼

On Windows, GPUs using the WDDM driver model share the device with the display, and the OS watchdog may kill long-running kernels. Check gpuDevice properties DriverModel and KernelExecutionTimeout, then increase the TDR timeout, close other GPU apps, or split computation into smaller chunks.

Does a "requires a newer graphics driver" message mean my GPU is unsupported?▼

Not necessarily. This message does not guarantee a newer driver exists or that the GPU is end-of-life. Consult the GPU Computing Requirements documentation page for your MATLAB release to confirm hardware support before updating drivers.

Related Skills

Explore More Software Engineering Skills→
dongg622dongg622

gemm-perf

Measure GPU matrix computation throughput with standardized TFLOPs or TOPs metrics.

Community
Intermediate
sparq-orgsparq-org
8

gpu-kernels

Benchmark GPU versus CPU performance for sparq columnar operations using wgpu compute kernels.

Official
Advanced
hsliuustc0106hsliuustc0106
84

vllm-omni-distributed

Distribute large models across GPUs or nodes with tensor and pipeline parallelism.

Community
Advanced
viniruggeriviniruggeri

get-available-resources

Detects CPU, GPU, memory, and disk resources and writes .claude_resources.json with strategy recommendations.

Community
Advanced