Lambda Labs, a GPU cloud service provider, has chosen VAST Data to manage AI training data across its extensive infrastructure of over 60,000 Nvidia GPU servers in cloud and co-location datacenters.
This agreement closely follows a similar partnership with CoreWeave last month. Like CoreWeave, Lambda Labs is experiencing growing demand for GPU services, particularly for large language model training.
Lambda Labs offers a scalable GPU cloud, co-location datacenters optimized for GPUs, and customer-owned Echelon hardware systems, all supported by an open source Lambda software stack.
Mitesh Agrawal, Lambda's head of cloud and COO, highlighted that "The VAST Data Platform allows Lambda customers with private cloud setups to seamlessly expand into Lambda's public cloud as workload demands evolve.
Moving forward, we aim to fully integrate VAST Data Platform features to maximize value for our customers' GPU cloud investments and data."
VAST Data's partnership with Lambda Labs leverages Nvidia SuperPod certification and its DataBase and DataSpace technologies, providing a unified global namespace for file and object storage across private and public cloud deployments.
This integration enables seamless data storage, retrieval, and processing for customers within a consistent environment.
Renen Hallak, CEO of VAST Data, expressed enthusiasm about the collaboration, emphasizing Lambda's leadership in AI cloud architecture. Together, they aim to deliver optimized cloud solutions tailored for AI workloads, enhancing efficiency in large language model (LLM) training and facilitating global collaboration.
VAST Data's DataSpace ensures that customer data remains accessible across Lambda's three environments, supporting hybrid cloud AI data management initiatives. As highlighted by VAST co-founder Jeff Denworth, several major Lambda customers are already poised to benefit from this integrated hybrid cloud approach.
Bootnote: Lambda Labs, founded by Michael Balaban and CEO Stephen Balaban in 2012, specializes in deep learning infrastructure, offering Nvidia GPU compute resources and co-location services.
The company has raised $112 million through multiple funding rounds, with plans for additional investment to expand its GPU server systems in response to growing demand, projecting revenues of $250 million for 2023 and $600 million for 2024 amidst the accelerating GenAI market surge.
