IBM and Together AI have entered a $240 million multi-year agreement to construct an AI inference cl...
The AMW Read
The deal updates Together AI's position in the inference infrastructure segment and signals dedicated compute demand, justifying the compute cross-ref.
IBM and Together AI have entered a $240 million multi-year agreement to construct an AI inference cluster on IBM Cloud, powered by Nvidia's HGX B300 systems. The deal highlights the growing demand for dedicated AI infrastructure and positions Together AI as a key cloud-native inference provider.
This partnership underscores the increasing shift toward specialized, high-performance inference infrastructure as enterprises deploy AI at scale. By leveraging IBM Cloud's enterprise-grade environment and Nvidia's latest B300 hardware, Together AI can offer more robust and efficient inference services to its customers, addressing the need for low-latency, cost-effective model deployment. This move also reflects the broader trend of neocloud providers partnering with established cloud and hardware vendors to meet enterprise demand.
For builders and enterprises, this means more options for deploying AI workloads with dedicated compute, potentially reducing reliance on hyperscaler default paths and enabling more tailored inference solutions. Investors should watch how such alliances strengthen Together AI's competitive position versus hyperscalers and other inference-focused providers. Per the AI Market Watch index, Together AI has raised $533.5M in total funding, and this deal could further accelerate its growth trajectory.
