IBM signed a $240 million multi-year agreement with Together AI to deploy a large-scale NVIDIA HGX B300 inference cluster. The new infrastructure will launch on IBM Cloud in the first quarter of 2027. This deployment marks the first cluster of its kind on IBM Cloud, specifically optimized for AI inference workloads.
IBM aims to capture the neocloud market by providing GPU-focused infrastructure to developers and enterprises. Together AI will utilize the cluster to support its open-source AI model inference business. The partnership strengthens IBM's competitive position in the global cloud and AI infrastructure sectors.