HARNESS IDLE
LAB POWER

Transform idle university lab PCs into your personal computing cluster. Run containerized ML training, simulations, and batch jobs on-demand.

Submit Your First JobSee How it Works

HOW IT WORKS

Choose Your Lab#01

Select from available university computing labs with idle high-spec PCs (32-64GB RAM, multi-core CPUs).

Configure Your Job#02

Specify your Docker container, resource requirements (CPU, memory, replicas), and submit to the job queue.

Run Containerized Workloads#03

Your job is scheduled across available nodes with gang scheduling, real-time logs, and automatic retry on failures.

WHAT CAN YOU RUN?

LeaseGrid supports any containerized workload. Here are common use cases:

Machine Learning & AI

Train models, fine-tune LLMs, and run hyperparameter searches across distributed nodes.

  • • PyTorch DDP training
  • • TensorFlow distributed
  • • Batch inference pipelines

Scientific Computing

Run simulations, analyze data, and process complex computational workflows.

  • • Monte Carlo simulations
  • • Molecular dynamics
  • • Climate modeling

Data Processing

Execute ETL pipelines, batch transformations, and large-scale data analysis.

  • • Pandas/Spark jobs
  • • Image/video processing
  • • Data transformation

CONTAINERIZED WORKLOAD EXAMPLES

python:3.11

PyTorch/TensorFlow training

jupyter/scipy-notebook

Scientific notebooks

nvidia/cuda:12.0-base

GPU-accelerated workloads

Custom Docker images

Your own ML pipelines

ABOUT THE CREATOR

Ebrahim Mamawala

Hello, I'm Ebrahim.

I built LeaseGrid to solve a universal problem: the high cost of cloud computing for batch workloads.

LeaseGrid transforms idle university lab PCs—sitting unused 75% of the time—into a Kubernetes-orchestrated batch computing cluster with fair-share scheduling and zero-trust security.

Whether you're a researcher, developer, startup, or enterprise—LeaseGrid helps you run compute-intensive workloads on underutilized infrastructure.

READY TO RUN YOUR FIRST JOB?

Get Started Now