Welcome to TensorDock

Welcome to TensorDock! Thanks for joining us. We're here to make your high-performance computing cloud deployments as cheap as possible.

Our Websites

  • Marketing Website: The static website where we market our products
  • Console: This is the dashboard where you can manage your deployments
  • API Documentation: This is our Postman site where you can view information about our API
  • Docs: This is the site you are on right now, where you can view non-API-related documentation

What is TensorDock?

Cloud high-performance computing, made cheap

Operating GPUs and controlled costs is in our DNA.
Before TensorDock was born, the team was working on another project, Dash Cloud. We had a slightly crazy plan to run our own DC in Montreal. We were working on a lease to use half of an industrial building to house a few hundred GPUs. We had even lined up someone who lives in Montreal to become the onsite technician. Before we did this though, we decided to try running our location in our home state of Massachusetts. We've played around with open-air chassis, modified crypto mining rig frames, natural airflow cooling (no fans = less power!), etc. We even considered getting 3M Novec for immersion cooling. Today, we run two of our own locations in the Greater Boston Area.

Integrate one platform, deploy anywhere

We partner with cloud providers around the world.
In addition to our own hardware, we're working on partnering with some of the top local providers around the world to enable you to deploy cloud GPUs at a dozen locations through our platform. Instead of needing to write 10 API integrations, one for each provider you integrate onto your deployment stack, you can just deploy through our platform. The prices you pay on TensorDock are just as low, if not lower than market rates. This is because with our scale, cloud providers give us discounts. So, buying compute through us saves you money. And, you develop one API integration (with us), instead of 10.
Yes, we say we "partner" with other cloud providers. Not exactly resell.
Our partners' success is our success, and our success is our partners'. Example: In Singapore, we guaranteed the scale for our local partner to negotiate lower colocation rates at Equinix SG1, enabling them to move from Ascenix without paying a premium. The servers in Singapore that we have access to are jointly owned.
Not a marketplace.
We control pricing, and we ensure quality through meaningful partnerships. There is a tendency for marketplaces to incentivize operators to reduce their prices. This comes at the cost of power redundancy, data storage quality, network quality, etc. With TensorDock, the servers we operate in Singapore jointly with a local provider are very much the same as the ones we operate from Boston.
When you deploy a server on TensorDock, you can sleep well, knowing that you'll either be placed on hardware that we fully own or with a supplier that we work very closely with, potentially with shared ownership over the hardware itself.

Unbeatable Stock

You don't have to worry about stock on TensorDock.
We might run out of our own GPUs, but because we partner with different cloud providers, we always have additional stock. Just try a different location.

High Caliber Team

We hire the self-taught, the unestablished, the fresh-out-of-school new grads. We don't have technical coding interviews (everyone can code well. Can't you use Stack Overflow on the job?). Instead, we show the candidate an excerpt of our codebase and ask them what would you do better? What products do you want to launch at TensorDock? Depending on the passion shown to our mission of democratizing high-performance computing, we may hire them. And then, they implement whatever they imagine. Water cooling? Try it! CPU instances? Why not? It's this culture that encourages fresh perspectives that led us to slash hardware costs by 80% without compromising quality by building our own servers instead of purchasing from OEMs. This strikes at the core of our culture: we innovate, and because we innovate, we stay ahead of the pack.

Real High-Performance Computing

TensorDock is not your regular cloud provider. We are your high-performance computing (HPC) provider, here to give unparalleled computing scale at industry-leading prices.
We provide GPUs and CPUs at market-leading prices from 5 locations throughout the United States. We're a team of passionate engineers, here to simplify high-performance computing.
Read more about our story, our culture, and our data centers.

Get Started

If you will be running an ML workload, see our machine learning guide.
If you will be cloud gaming, see our cloud gaming via Parsec guide.


We're always available at [email protected]. If you have a favorite team member, you can reach out to them directly, as all of our team members help with support!
Last modified 1mo ago