The news
The Nutanix Ryax acquisition, announced September 22, brings a French compute orchestration company into Nutanix (NTNX) as the hybrid cloud software maker tries to help enterprises run AI agents on their own infrastructure without leaving expensive chips idle.
Nutanix did not disclose a price and said the deal's financial impact is not material. Ryax, based in France, sells an AI-driven platform for orchestrating and managing compute. Its team will stay in France and keep developing the technology, which Nutanix plans to build into future releases of Nutanix Kubernetes Platform (NKP) and Nutanix Enterprise AI (NAI).
Two capabilities anchor the deal, according to Nutanix. Intelligent Resource Optimization aims to get more work out of the GPUs and CPUs customers already own. AI-Aware Smart Scheduling automatically places AI workloads on hardware based on performance and cost. Techzine, which covered the deal on September 23, reported that Ryax sizes resources for each execution and supports fractional GPU slicing so several containers can share one card. In Ryax's own tests, per-execution sizing cut node-hours by 62% on a 30-run deep-learning burst while finishing the work 5.7% faster, Techzine said.
Nutanix chief executive Rajiv Ramaswami said the purchase addresses challenges enterprises face in managing infrastructure for agentic AI, meaning AI systems that plan and carry out multi-step tasks. Ryax chief executive Andry Razafinjatovo said the combination would help organizations accelerate their agentic AI initiatives.
The deal lands as Nutanix courts customers rethinking VMware. On September 22, The Register reported on Gartner research predicting that by 2029, 55% of enterprises will start proofs of concept for alternative distributed hybrid infrastructure products to replace their VMware-based deployments, up from 25% in 2026, citing cost increases after the shift to per-core subscriptions under Broadcom (AVGO). The same report said Gartner placed Nutanix among market leaders but described its licensing as complex and less competitive.
The numbers
- Purchase price
- Not disclosed
- Node-hour reduction in Ryax's 30-run deep-learning test (Techzine)
- 62%
- Enterprises expected to start proofs of concept to replace VMware by 2029 (Gartner, via The Register)
- 55%, up from 25% in 2026
Why CEOs should care
For infrastructure leaders, GPU utilization is becoming a budget line as important as the GPUs themselves. If you run NKP or Nutanix Enterprise AI, ask Nutanix when Ryax's scheduling will ship, whether it will carry an extra license fee, and how it will work alongside schedulers you already use. Ask for utilization results on your own workloads; a 62% reduction in a vendor's test is a starting point, not a forecast.
CFOs approving AI hardware should ask what share of purchased GPU capacity is actually in use. Software that packs more jobs onto existing accelerators can defer new purchases, but only if teams measure utilization before and after. Tie future AI infrastructure requests to a utilization target and a review date.
Companies weighing a move off VMware should treat AI workload support as part of that evaluation rather than a separate project. If Gartner's forecast holds, many peers will be running trials of alternatives by 2029. Boards should ask whether the company's hybrid cloud plan accounts for licensing cost, migration risk and readiness to host AI agents on its own hardware.
The bigger picture
Enterprise infrastructure software vendors are shifting from selling platforms to promising efficiency on scarce, costly AI hardware. Nutanix wants customers to run agentic AI on premises, in public clouds, in neoclouds and on high-performance computing clusters, Techzine reported, with smarter scheduling as the connecting layer. Small tuck-in deals like this one, with undisclosed prices and immaterial financial impact, let vendors add specialized AI infrastructure skills quickly while competing for customers leaving VMware.
The Register's account of the Gartner research also suggests pressure across the whole category, not only on VMware. It said Gartner cited customer complaints about VMware's communication, commercial practices and support delays, and said AWS Local Zones carry a 15% to 35% price premium over their parent regions in expensive metros. Gartner clients also reported minimal flexibility in VMware negotiations. Still, The Register noted that rivals' progress in luring VMware customers is slow and said Broadcom's VMware unit will keep most of its customers for at least another three years, which points to a gradual shift rather than a quick exit.
What’s next
Nutanix has not given a release date for Ryax capabilities in NKP or Nutanix Enterprise AI. Watch for roadmap details and pricing in Nutanix's next earnings report and product updates, for whether the features are bundled into existing licenses or sold as an add-on, and for any further tuck-in acquisitions aimed at AI infrastructure efficiency.
What “Fact-checked” means
Fact-checking means testing a story’s facts against the evidence before it is published. This story went through at least two separate checks before this version was published.
- What we checked
- Its names, figures, dates, job titles, quotes and who said what were checked against the story’s sources, including its main source where it could be opened. The headline was checked for accuracy and overstatement.
- How
- A first check reviewed the whole story. If it passed, a second, skeptical check went back to the sources to look for mistakes in the most important facts. If a check flagged the story, it was edited to fix the problems found, and a separate re-check then reviewed the whole story again.
- Who
- The checks are made with our newsroom’s technology tools, as steps kept separate from the writing, under rules set by our editor, Hussein Mukhtar. A story the checks still flag is not published automatically; it is held for the editor, who decides whether it is fixed, published or dropped.
- If something is wrong
- “Fact-checked” does not mean error-free. If a material error is found after publication, we correct the story and add a note saying what changed. Report an error
Companies in this story
