Need a NVIDIA B300 cluster by January, at an unbeatable price? Call us.
The San Francisco Compute Company
We run large-scale, vetted GPU clusters. Need 800 H100s on a single InfiniBand fabric for an hour? You got it. 8 H200s dedicated for a year? Sure. Need to cancel? Just sell back what you don't need. You're never locked-in to a long-term contract.
Never pay to move your data. We run the clusters from UEFI on up and support you just like a traditional cloud.
Average prices for H100 nodes
Pricing shown per gpu/hour based on selected duration and cluster size. Need a custom configuration? Use our CLI or go directly to the buy page.
A simple but powerful CLI Read docs
# Buy 32 nodes (256 H100s) for 3 days at market price $ sf nodes create -n 32 --zone landsend -d 3d # Maintain 2 nodes (16 H100s) when the price is below $2.00/GPU/hr $ sf nodes create rl-node-1 rl-node-2 --auto --zone landsend -p 16.00
Used to train hundreds of large scale models
"Very competitive low prices, an abundant amount of H100s for our projects, and great and responsive customer service."
"We trust SF Compute to be our primary cloud for training
and inference workloads. The hardware spec is great and the team is amazing to work with."
Michael Royzen, CEO –Phind
"SF Compute has made reserving compute radically easier and simpler for us, with its transparency in pricing and availability of H100s. We are impressed by the SF Compute team's fast and reliable technical support
during training."
FAQ
Is it bare metal? Containers? VMs? Kubernetes?
When you buy on the website/CLI, you'll get virtual machines. If you'd like a bare-metal or Kubernetes cluster, contact us.
How fast do VMs spin up?
About 5 minutes.
Are nodes fully-interconnected with InfiniBand?
Not yet, but we'll support InfiniBand in VMs Q1 2026.
Will the nodes go down?
Yes. Hardware failure rates are much higher on GPU clusters than on web servers. At certain scales, they're guaranteed, so we've designed for failure. We have strict hardware requirements and have seen just about everything that can go wrong. Unlike other providers, we refund for failed nodes and can repack your nodes to ones with healthy hardware.
What support do you provide?
Large clusters get a dedicated Slack channel with our engineering team and emergency phone escalation. Our team has managed $1B+ of hardware and worked on several of the world's supercomputers.