Lambda Inc., a startup that operates an on-demand public cloud for training and deploying artificial intelligence models, today announced it has raised $320 million in a new funding round increasing the company’s valuation to $1.5 billion.

The Series C round was led by Thomas Tull’s US Innovative Technology with participation from new investors B Capital, SK Telecom and existing investors Crescent Cove, Mercato Partners, 1517 Fund, Bloomberg Beta and Gradient Ventures, among others.

Founded in 2012, Lambda provides cloud, on-premises and consulting services to companies looking to train AI models. The company began selling AI infrastructure in 2017 when the modern software architecture of the AI transformer, the basis behind large language models such as OpenAI’s ChatGPT, was outlined. Back then, the company was used primarily by academic researchers as a platform for image recognition, speech generation and natural language processing.

Now, Lambda’s platform delivers access to large clusters of Nvidia Corp.’s graphics processing units that allow customers to train AI models and deploy them. Within the startup’s fleet is Nvidia’s flagship H100 data center chip, which is built on the Hopper architecture, capable of training even the most complex AI models, enabling extremely powerful chatbots and image generating AI.

“AI is fundamentally restructuring science, commerce, and industry. Over the next 10 years, every human endeavor will be augmented by the integration of LLMs and generative AI,” said Lambda co-founder and Chief Executive Stephen Balaban. “This AI rollout is going to require a lot of GPUs. This latest financing supports our mission to make GPU compute as ubiquitous as electricity.”

The company said that it would use this new funding to scale out its Nvidia chip offerings to grow its GPU cloud to ensure access to AI developers and engineers and build more network apparatus using Nvidia’s Quantum-2 Infiniband networking capabilities.

According to the company, Lambda’s private cloud business serves over 5,000 customers across numerous industries including manufacturing, healthcare, pharmaceuticals, financial services and U.S. government interests. Big customers and research institutions that rely on the AI cloud include the scalable AI platform Anyscale Inc., Japanese technology conglomerate Rakuten Group Inc. and The AI Institute.

Aside from its cloud offering, Lambda also offers on-premises hardware with AI-optimized appliances and personal computers. The company sells a family of servers that contain Nvidia H100 GPU clusters that companies can buy for their own data centers, or have them colocated at Lambda’s own facilities. For AI engineers and workers, the company also sells two different desktop PCs with Nvidia GPUs and a Tensorbook laptop designed to handle high-performance deep learning workloads.

Image: Pixabay