skills.rest
dongzhuoyao
dongzhuoyaoCOMMUNITY
Explore all skills by this author →
Website𝕏@vtaohuGitHub

sue-cluster-info

Query live Slurm job limits and GPU hardware specs for LUMI and Snellius clusters.

Updated Aug 27, 2026
Search Tags:#hpc#slurm#snellius#lumi#gpu cluster#scale-up#job limits
One-click install
npx skills add https://github.com/dongzhuoyao/deepresearch --skill sue-cluster-info
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: sue-cluster-info
Source: https://github.com/dongzhuoyao/deepresearch/tree/main/.codex/skills/sue-cluster-info
Command: npx skills add https://github.com/dongzhuoyao/deepresearch --skill sue-cluster-info

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Manually gathering accurate, up-to-date GPU cluster hardware specifications and account-specific job limits for LUMI and Snellius supercomputers is time-consuming and error-prone, often leading to failed scale-up experiment submissions or quota overruns due to stale documentation.

Core Features & Use Cases

  • Live Job Limit Queries: Pulls real-time Slurm account-specific running and pending job caps, active job counts, and remaining headroom instead of relying on outdated published defaults.
  • Static Hardware Summaries: Provides comparable, side-by-side hardware details for LUMI (ROCm/MI250X) and Snellius (CUDA/A100/H100) clusters including GPU memory per node, node counts, and partition rules.
  • Use Case: Before submitting a large-scale ML training run, use this skill to quickly compare LUMI and Snellius capacity, confirm your account has sufficient job headroom, and select the correct partition to avoid submission failures.

Quick Start

Use the sue-cluster-info skill to get a summary of LUMI and Snellius GPU cluster hardware and your account's live job limits for scale-up experiment planning.

Frequently Asked Questions about sue-cluster-info

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

FAQPage Schema
How do I check live Slurm job limits and GPU capacity on LUMI or Snellius before submitting ML scale-up experiments?▼

To check live Slurm job limits and GPU capacity for scale-up experiments, query real-time account-specific running and pending job caps, active job counts, and remaining headroom alongside static hardware specifications for LUMI and Snellius clusters.

What are the hardware differences between LUMI and Snellius GPU clusters for HPC workloads?▼

LUMI utilizes ROCm-based MI250X GPUs, while Snellius uses CUDA-based A100 and H100 GPUs. Comparing these hardware specifications side-by-side helps select the correct partition and compatible software stack for your specific ML or HPC workload.

Why do my scale-up experiment submissions fail due to Slurm job quotas on GPU clusters?▼

Scale-up experiment submissions often fail due to relying on stale documentation for GPU cluster job limits. Querying live Slurm account-specific job caps and active counts prevents exceeding job quotas and avoids submission failures during large-scale ML training runs.

Can I compare node memory and partition rules for LUMI and Snellius before a dryrun?▼

Yes, you can compare GPU memory per node, node counts, and partition rules for both LUMI and Snellius clusters side-by-side. This ensures your dryrun and fullrun submissions select compatible hardware without violating partition constraints.

How do I confirm my account has sufficient job headroom on Snellius for a fullrun submission?▼

To confirm sufficient job headroom on Snellius for a fullrun submission, query live Slurm account-specific running and pending job caps and active job counts. This provides accurate, account-scoped capacity metrics to ensure safe submission without quota overruns.

Related Skills

Explore More Education & Research Skills→
niznik-devniznik-dev
16

Run Experiment

Orchestrate SLURM fine-tuning jobs with torchtune and evaluate with inspect-ai.

Community
Advanced
caiokfcaiokf

gob-background-jobs

Manage parallel background commands with gob job IDs and output capture.

Community
Advanced
dongzhuoyaodongzhuoyao
10

lumi-supercomputer

Configure Slurm and ROCm environments for PyTorch jobs on LUMI.

Community
Advanced
umich-foreseerumich-foreseer
16

slurm-job

Generate Slurm sbatch scripts with partition, account, GPU, and validation directives.

Official
Intermediate