Dell Technologies and Imbue Collaborate on $150 Million High-Performance Computing Cluster

Dell Technologies and Imbue, an independent AI research company, have announced a partnership to develop a new high-performance computing cluster. The cluster will be used to train foundation models optimized for reasoning. Imbue is known for developing its own foundation models and training them to have advanced reasoning capabilities. The goal is to create practical tools for building AI agents that can assist workers in various fields.

Imbue is already using the cluster, which is powered by Dell PowerEdge XE9680 servers with NVIDIA H100 Tensor Core GPUs, to train AI models and develop prototype agents. These agents can correct bugs in code and analyze lengthy documents. The long-term plan is to design AI agents that are more capable and trustworthy, eliminating the need for constant supervision from users.

The system, designed by Imbue and Dell, includes smaller clusters for rapid experimentation on novel model architectures. It also features rapid networking into a large cluster to enable efficient training of large-scale foundation models.

"Dell Technologies has helped us deploy a custom cluster much more quickly than other providers could have," said Josh Albrecht, CTO of Imbue. "Dell has been an invaluable collaborator as we pursue our work to create AI systems with much stronger reasoning abilities."

The Dell systems used in the cluster are built for extreme acceleration in AI, machine learning, and deep learning training. They are equipped with high GPU memory, bandwidth, and security features. The PowerEdge servers' Smart Cooling features ensure great performance while reducing the data center's carbon footprint.

The system is managed by Voltage Park, a cloud compute provider specializing in machine learning solutions.