A GPU cloud for molecular dynamics.
Run MD simulations and generate data for AI molecular models on thousands of GPUs starting at $0.015 per hour. Our open-source batch binary handles checkpointing and auto-scaling so interruption never costs you a trajectory.
Simulation sweeps are made for a distributed network.
Ensemble runs, replica exchange, and free-energy calculations parallelize across independent GPUs with no interconnect requirement. Consumer RTX cards deliver strong single-precision throughput per dollar, and checkpointing to storage makes preemption harmless. For very large systems that need 96 GB on one device, Salad Dedicated adds reserved RTX PRO 6000 Blackwell capacity.
- GROMACS, OpenMM, and custom engines in your own container
- Long-running task patterns with RabbitMQ, SQS, or Kelpie
- Automatic reallocation and resume from checkpoint
"If you want access to 1000s of GPUs, you can get them on SaladCloud for better cost-efficiency. Salad is also really customer friendly, something you cannot get from a larger cloud provider as a startup."
Zachary Lawrence · CEO, Klyne.ai · AI drug discoveryCase study →