By TradingView
Publication Date: 2026-08-17 12:18:00
IBM IBM is deepening its push into artificial-intelligence infrastructure with a $240 million multi-year deal to build a massive Nvidia-powered inference cluster for Together AI. The agreement gives IBM Cloud a high-profile AI workload while strengthening Nvidia’s position in the fast-growing inference market, where companies are shifting from training models to running them at scale.
IBM generates revenue across software, consulting and infrastructure, with its cloud business supporting companies running workloads across hybrid environments. AI has become increasingly central to that strategy as IBM looks to combine infrastructure with higher-margin software and enterprise services.
Under the agreement, IBM will deploy a large cluster of Nvidia HGX B300 systems on IBM Cloud, with availability expected in the first quarter of 2027. Together AI will use the capacity to provide open-source model inference to enterprise customers.
The deployment will be IBM Cloud’s first dedicated large-scale inference cluster built around HGX B300 systems and Nvidia Spectrum-X Ethernet networking. Nvidia 9NVDA) says the configuration can deliver 30 times more AI factory output than previous generations.
The scale of Together AI’s existing workloads makes the deal noteworthy. The company says its inference platform already processes about 400 trillion tokens monthly, while its broader platform covers training, fine-tuning and agentic AI workloads.
For IBM, the contract arrives as its broader…