Compute
Find reliable compute, from a single GPU to largest clusters.
Instant access to 1–256 GPUs on demand, or reserved clusters sourced from 50+ providers — unified in a single platform with InfiniBand, SLURM and Kubernetes.
| GPU | Configuration | Interconnect | On demand | Action |
|---|---|---|---|---|
| B300 | 288 GB VRAM · 480 GB RAM · 48 vCPUs | NVLink 5 · InfiniBand | $4.99/HR | Launch → |
| B200 | 192 GB VRAM · 384 GB RAM · 32 vCPUs | NVLink 5 · InfiniBand | $3.49/HR | Launch → |
| H200 | 141 GB VRAM · 182 GB RAM · 44 vCPUs | NVLink 4 · InfiniBand | $3.14/HR | Launch → |
| H100 | 80 GB VRAM · 185 GB RAM · 32 vCPUs | NVLink 4 · InfiniBand | $2.43/HRSpot $0.94/HR | Launch → |
Per GPU-hour · on-demand list rates. Reserved and multi-node pricing quoted per request.
On demand
Instant access to 1-256 GPUs. Use your GPUs across clouds in a single platform.
- 1.1
SLURM, K8s Orchestration
Orchestrate dynamic workloads with enterprise-grade scheduling and container automation.
- 1.2
Infiniband Networking
Scale distributed training with high-bandwidth interconnects across nodes.
- 1.3
Grafana Monitoring Dashboards
Visualize metrics in real time with customizable dashboards for full system observability.
Liquid Reserved Clusters
Request large-scale clusters from 50+ providers. Sell-back idle GPUs to our spot market.
- 2.1
Get quotes from 50+ datacenters within 24 hours
One request, parallel bids for options, from H100, H200, to B200, B300, GB300 NVL72
- 2.2
Re-sell idle GPUs back to our spot market
Resell idle nodes on our spot market with no manual ops. Reclaim capacity instantly when you need it.
- 2.3
Direct assistance from our research and infra engineering team
Dedicated solutions engineer from cluster bring-up through steady-state
FAQ
Questions, answered.
How quickly can I get GPUs?
On-demand capacity of 1–256 GPUs is available instantly. For reserved clusters, one request collects parallel bids from 50+ datacenters, with quotes within 24 hours.
Which orchestration do you support?
SLURM and Kubernetes, with InfiniBand networking for multi-node jobs and Grafana dashboards for real-time observability.
What happens to idle reserved capacity?
Resell idle nodes on our spot market with no manual ops, and reclaim capacity instantly when you need it.
[ Ready? ]
Your model should get smarter while you sleep.
Engineers, at full stride. Point PI at your checkpoint — first improvement in about 48 hours.