TL;DR — Key Takeaways
– Nutanix has acquired French software developer Ryax Technologies to improve GPU utilization and AI workload orchestration across hybrid cloud environments.
– Ryax’s technology will be integrated into future releases of Nutanix Kubernetes Platform and Nutanix Enterprise AI to automate resource allocation and workload scheduling.
– The acquisition aims to reduce the cost of deploying agentic AI by helping enterprises maximize existing CPU and GPU capacity.
Nutanix has acquired French software developer Ryax Technologies to improve GPU utilization and simplify the use of agentic AI across hybrid cloud environments. The acquisition will expand Nutanix’s ability to manage AI workloads across private data centers and public clouds.
Nutanix plans to incorporate Ryax’s compute orchestration technology into future releases of Nutanix Kubernetes Platform and Nutanix Enterprise AI. The integration will allow customers to automatically select computing resources based on workload requirements and costs.
This acquisition addresses an efficiency challenge: expensive GPUs are often underutilized even as businesses struggle to obtain sufficient computing capacity. Many organizations operate AI infrastructure across mixed environments, including hyperscalers, specialized GPU cloud providers known as neoclouds, and their own data centers.
Managing all these resources requires considerable manual configuration, particularly when moving AI applications from development into production. Ryax’s technology is designed to automate much of this work.
For Nutanix, the acquisition strengthens its hybrid cloud strategy by extending infrastructure management into AI workload optimization. Rather than requiring enterprises to purchase additional GPUs for every new AI application, the combined platform aims to help customers extract more from existing hardware. In particular, the technology is designed to lower the costs of deploying agentic AI, which can consume large levels of computing resources. Nutanix is focusing on intelligent resource optimization and AI-aware smart scheduling.
Resource optimization will help enterprises improve the efficiency of available CPUs and GPUs. Smart scheduling will automatically direct AI applications toward hardware that meets performance requirements while controlling costs.
For Nutanix Enterprise AI, the planned integration includes a scheduling layer capable of distributing training, inference and batch workloads across NVIDIA and AMD hardware, cloud environments and HPC clusters.
Working Across Distributed Environments
Founded in Lyon, France, in 2017 by CEO Andry Razafinjatovo and CTO Yiannis Georgiou, Ryax developed a platform that manages containerized AI applications across distributed computing infrastructure.
The software handles application packaging, deployment, scheduling, scaling and monitoring. It supports Kubernetes environments and Slurm clusters used for HPC, allowing developers to deploy workloads across different infrastructure without rewriting application code.
A key capability of the platform is intelligent GPU allocation. Ryax supports fractional GPUs, enabling multiple AI applications to share a single processor rather than requiring each workload to reserve an entire GPU.
The platform also uses historical performance data to determine the computing resources required for individual jobs. If an application encounters insufficient GPU memory, the software can adjust its allocation and restart execution.
Ryax reported major efficiency gains in its own testing. In one test involving 30 deep-learning executions, automated resource allocation reduced node-hours by 62% while completing the work 5.7% faster. Another test used NVIDIA’s Multi-Instance GPU technology to run four workloads simultaneously on a single H100 GPU, reducing the cost per execution by 52%.
The Ryax team will join Nutanix in France. The financial terms of the deal were not disclosed.