The cloud computing market is enormous — and it’s only expected to grow. According to a report jointly published by CenturyLink and Statista, it’ll be worth $411 billion by 2020.
With incumbents like Amazon Web Services, Microsoft’s Azure, and Google Cloud Platform competing for a slice of the pie, it’s not easy to gain a foothold. But New York City startup and Y Combinator graduate Paperspace found early success in a niche: artificial intelligence (AI) and machine learning. The Brooklyn-based infrastructure-as-a-service (IaaS) provider now powers tens of thousands of organizations with virtual machines for performance-intensive design, visualization, and AI apps, and it has its eyes set on expansion.
To that end, Paperspace today announced it has raised $13 million in a Series A funding round led by Battery Ventures, SineWave Ventures, Intel Capital, and Sorenson Ventures, as well as an existing investor — Reddit cofounder Alexis Ohanian’s Initialized Capital.
“AI is the key to the next generation of scientific breakthroughs and has the potential to impact businesses as the industrial revolution did for manufacturing, yet AI today is too complex, too exclusive, and too costly to reach the mainstream,” CEO and University of Michigan graduate Dillon Erb said. “Only large companies have the sufficient resources — infrastructure, tooling, and talent — to efficiently develop and deploy models into production.”
Setting up a Paperspace server is dead simple. After signing up for service, users are prompted to configure their virtual machine or select from one of several template collections. Once they’ve settled on the specifications, they’re off to the races and can launch and interact with their instance from within a browser window.
“The experience we deliver is a latency of about 10 milliseconds,” Erb told VentureBeat in a previous interview. “It’s amazing, and it’s a test to see whether a machine is running locally or on our technology.”
Earlier this year, Paperspace launched Gradient, a suite of enterprise tools designed to make developing cloud AI solutions “as simple as building a modern web service,” in Erb’s words. Among the headliners are a graphics processing unit (GPU) job runner, one-click Jupyter notebooks, and an integration that lets customers run any routine on a GPU-accelerated cloud by adding a single line of code.
Gradient supports popular machine learning frameworks, including Google’s TensorFlow, Facebook’s Caffe2, and PyTorch, and it includes enterprise-tailored capabilities such as data integration, user management, access control, and the ability to orchestrate across multiple cloud providers. That’s all in addition to features like the ability to run multiple monitors in a web browser, integration with virtual private networks, deployment in ActiveDirectory environments, 1-click backups, real-time monitoring, custom templates, and shared drives.
“The key to unlocking AI is to give every developer access to these same resources. That’s our core mission with Gradient — to provide a powerful yet intuitive machine learning platform that any developer can leverage, whether you’re an individual AI practitioner or a large enterprise.”
Paperspace, which was founded in 2014, has raised a total of $19 million to date. (An early investor was Jeff Carr, one of the cofounders of cloud hosting company DigitalOcean.) The company runs its own datacenters with custom-configured GPUs and charges $8 a month for access to an entry-level Gradient machine, plus a bit extra for storage.