IBM, Together AI ink $240 million deal for Nvidia-powered AI inference cluster
IBM and Together AI signed a multi-year $240 million deal to build a large AI inference cluster on IBM Cloud using Nvidia HGX B300 systems and Spectrum-X networking, the companies said. The cluster will serve open-source models as firms seek lower AI costs and address cybersecurity concerns. Together AI’s platform supports models including DeepSeek and Kimi.
How this was made
The 30-second read
Why it matters
A named, multi-year IBM Cloud deployment using Nvidia Blackwell-era HGX B300 systems is a concrete signal of ongoing enterprise AI infrastructure buildout, with potential positive read-through for both IBM Cloud services and Nvidia’s inference hardware ecosystem.
Market read
Traders can treat this as a fresh, deal-based datapoint for AI inference infrastructure demand, with direct ecosystem implications for IBM Cloud and Nvidia hardware.
What to watch
Key missing details include customer concentration, contract duration economics, and whether Together AI’s platform adoption translates into sustained inference volumes on IBM Cloud.
Background
Inference demand has become a major driver of AI infrastructure spending as enterprises seek lower-cost, open-model deployments.
Ticker impact
IBM signed a $240 million multi-year agreement with Together AI to build an Nvidia-powered inference cluster on IBM Cloud.
Moderate positive bias for IBM shares as it signals continued enterprise AI spend capture.
The article provides deal size, duration, and specific infrastructure components, but does not quantify margins, customer count, or timing of revenue recognition.
The cluster will use Nvidia HGX B300 systems and Spectrum-X Ethernet, with Blackwell chips positioned as optimized for AI inference.
Mild positive read-through for NVDA, likely more sentiment than immediate financial impact.
The article names Nvidia systems but does not disclose Nvidia’s commercial terms, volume, or whether Nvidia is a contracted supplier with disclosed revenue impact.
Market effects
Reinforces the shift toward inference-optimized deployments on cloud platforms using Nvidia Blackwell-era systems.
Limited direct regional impact; enterprise AI infrastructure demand is global but deal is operationally tied to IBM Cloud.
Supports the broader AI capex/inference buildout narrative and open-model adoption driving compute demand.
Counterpoint
The deal may be more about platform positioning than near-term earnings impact if revenue recognition is back-end loaded or margins are thin.
Key entities
- public_companyIBM
Signed a $240 million multi-year agreement to host an Nvidia-powered inference cluster on IBM Cloud.
- startupTogether AI
Provides the platform to train and run open-model AI workloads and partners on the inference cluster.
- public_companyNvidia
Supplies HGX B300 systems and Spectrum-X networking gear used in the cluster.
