Announcing SFC Private Beta

The San Francisco Compute Company

We run affordable pre-training clusters for startups, grad students, research labs, other cloud providers, and large enterprises.

We sell compute by the hour, not by the year. Need 768 H100s for a week? You got it. Want to do 32 runs on 8 H100s for 2 hours each? Sure. You can flexibly burst to large portions of the cluster so you can scale up your model without having to sign an expensive long-term contract.

Get a cluster at $2.71/gpu/hr
*Prices are from the sfcompute private beta and may increase over the coming months.
Average Prices
Price ChartOct 13Oct 20Oct 27Nov 03$0.80$1.00$1.20$1.40$1.60$1.80$2.00$2.20$2.40$2.60$2.80$3.00Price
*Prices are from the sfcompute private beta and don’t represent normal market conditions.

A simple but powerful CLI Read docs

~/training
# Buy 256 H100s for 3 days at a max price of $2.30/GPU/hr 
$ sf buy -n 256 -d 3d -p 2.3

# Sell part of a contract for $2.10/GPU/hr 
$ sf sell -c <contract> -s tomorrow -d 1d -p 2.1

# SSH into a running VM 
$ sf ssh <instance>

Used to train hundreds of large scale models

PrincetonPlay.htPhindliquid
"Very competitive low prices, an abundant amount of H100s for our projects, and great and responsive customer service"
Tianyu Gao, PhD Student profile pictureTianyu Gao, PhD Student –Princeton Language and Intelligence
"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 Phind profile pictureMichael 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

Are the nodes fully-interconnected with InfiniBand?

Yes. If you buy 10 nodes, you’ll get a 10 node cluster with 3.2tb/s IB.

Is it bare metal? Containers? VMs?

It’s highly optimized VMs that perform like bare metal. One VM per host at the moment. You can SSH into each node and setup services like Kubernetes, Slurm, or your own orchestration stack.

Is there shared storage?

Yes. Right now it’s CEPH, but we’ll be rolling out other options soon.

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.

How does support work?

If you buy a big enough cluster, we’ll setup a shared slack channel with our engineering team. In the case of emergencies, we give you a phone number that escalates pages. Our team has managed over a billion dollars of hardware and worked on several of the world’s supercomputers.

Looking for a larger cluster?

We offer competitive pricing for large reservations of 512 - 4096 GPUs, on contracts with volume pricing that you can exit. Contact sales and let us know what you’re looking for.

Who are we?

We have been written about in The New York Times, The Verge, and on Hacker News. We have a Twitter X account, @sfcompute. You can find us in-person in San Francisco, the great city of progress.
© 2024 San Francisco Compute
PrivacyTerms
made in SF