Run GPU molecular simulations in minutes without a cluster

Most computational chemists outside a dedicated HPC team face the same bottleneck: waiting for cluster access or spending days on infrastructure setup before running a single simulation. This session shows a faster path: GPU-accelerated OpenMM simulations running as containerized batch jobs on Nebius, submitted directly from Python with automatic GPU provisioning and zero cluster configuration.

We’ll be joined by our partner Quantori, who will demonstrate how they use this approach in practice to run large-scale molecular simulations without managing infrastructure.

We’ll run a real protein-ligand simulation live, launch a parallel parameter sweep from a single script, and show you how results land in object storage automatically. No YAML, no Kubernetes, no waiting.

What you’ll see

• A containerized OpenMM simulation submitted and running in under 2 minutes from a Python script
• A parallel sweep across force field or temperature parameters launched with a loop, no additional setup
• Trajectory and log outputs automatically written to object storage on completion

Who should attend
Computational chemists and bioinformatics researchers running MD simulations, conformational sampling, or parameter sweeps who want GPU access without managing infrastructure.