Hybrid cloud software provider Nutanix has acquired France-based AI orchestration company Ryax Technologies for planned use in future releases of its platforms.
The financial terms of the deal were not disclosed.
Go deeper with GlobalData
Access deeper industry intelligence
Experience unmatched clarity with a single platform that combines unique data, AI, and human expertise.
Nutanix plans to incorporate Ryax’s graphics processing unit (GPU) utilisation and smart scheduling capabilities into future releases of Nutanix Kubernetes Platform (NKP) and Nutanix Enterprise AI (NAI). The company said the deal supports its strategy to enable enterprises to build, run and govern agentic AI across infrastructure environments.
Ryax develops an AI-driven compute orchestration and management platform and is based in France. Its technology is designed to automate the infrastructure, data, hardware and code involved in AI projects, reducing the integration and manual configuration work required to move deployments from experimentation into production.
Nutanix said demand for efficient AI inference workloads is increasing, while GPU shortages and fragmented infrastructure are forcing some organisations to seek capacity across private data centres, hyperscalers and neoclouds. It said a unified hybrid-cloud operating model is needed to simplify this complexity, improve hardware use and support economically viable deployments.
Nutanix CEO Rajiv Ramaswami said: “Agentic AI requires a flexible architecture, simple operations and the ability to use infrastructure far more efficiently.
“Our acquisition of Ryax helps to directly answer the challenges enterprises face today with agentic AI infrastructure management. The acquisition marks an exciting milestone in our commitment to simplifying enterprise AI at scale.”
The planned integration focuses on intelligent resource optimisation and AI-aware smart scheduling. The first is intended to improve the use of available GPUs and CPUs, while the second would automatically place AI workloads on cost-effective hardware to support performance and budget management.
Nutanix also outlined automated placement of jobs across multi-cluster, public-cloud and high-performance computing environments. Other planned functions include telemetry-driven, per-execution resource sizing to replace static allocations, and a single native layer for deploying, operating and scaling production AI stacks.
Future NKP capabilities are expected to include right-sizing and automated recovery from out-of-memory errors, using historical resource profiles to set GPU allocations, increase VRAM and retry jobs. Nutanix also plans fine-grained GPU bin-packing, allowing fractional GPU slices to be assigned to multiple containerised workloads on shared hardware.
Serverless GPU capacity would be allocated only during active compute cycles and released when the work is completed, rather than remaining idle. Nutanix said this could make capacity available for other workloads, including training runs.
Ryax CEO Andry Razafinjatovo said: “The Ryax platform was built to give enterprises seamless execution across hybrid infrastructure, like private data centers, hyperscalers and neoclouds, handling the backend complexity so teams can deploy, scale and manage AI workflows from day one.
“By combining our technology and talented team with Nutanix and its hybrid cloud platform, we will further empower organisations to accelerate their agentic AI initiatives seamlessly, wherever their infrastructure lives.”
