.png)
Spectro Cloud is announcing that PaletteAI has completed NVIDIA AI Cloud Ready validation.
If you operate a neocloud, a sovereign AI cloud, or an AI-as-a-service business, getting AI factories into production quickly is only the beginning. You also need the reliability, consistency, and scale to operate them as a service. PaletteAI provides that software foundation and has now been validated against NVIDIA AI Cloud Ready functional requirements for production AI cloud infrastructure software.
“Enterprises and governments are standing up production AI factories at speed, and the NVIDIA AI Cloud Ready Initiative identifies the software designed to support production-scale AI cloud environments. Spectro Cloud PaletteAI has now cleared that bar.” —Warren Barkley, Vice President, Product Management, NVIDIA
NVIDIA AI Cloud Ready Initiative
The NVIDIA AI Cloud Ready Initiative provides a technical validation framework for infrastructure software deployed on NVIDIA Cloud Partners (NCPs). Under the program, solutions are assessed against the NCP Software Reference Guide, a reference guide for the capabilities needed across AI cloud infrastructure and platform layers.
PaletteAI is among the ISV software solutions to complete this validation process. To complete this validation, the required infrastructure and platform capabilities, such as GPU scheduling, tenant isolation, inference workloads, and lifecycle automation, have to be present and working under representative AI workloads. PaletteAI has to integrate cleanly with the NVIDIA accelerated computing stack and behave predictably across platforms under production AI load. Published validation results document the stack version, test criteria, results, and configuration.
Completing NVIDIA AI Cloud Ready validation comes as the latest milestone in our work with NVIDIA. Spectro Cloud has also been previously recognized in the Enterprise AI Factory validated design for Blackwell and the AI Factory for Government design. Our broader relationship with NVIDIA spans the NVIDIA Partner Network, where Spectro Cloud is a Preferred Partner, the NVIDIA Jetson™ Ecosystem, and the NVIDIA Inception program, from which we have graduated.
Multi-tenant Kubernetes on AI infrastructure
PaletteAI’s validation covers what an AI cloud operator needs to deliver Kubernetes as a service on GPU infrastructure.
PaletteAI isolates tenants at the Kubernetes layer: tenant-scoped operations, project boundaries, RBAC wired into the identity systems your customers already use, and quotas enforced across shared GPU pools. Sovereign workloads that share infrastructure with commercial tenants operate in their own control planes with their own policies.
The validation also covers the AI workloads operators run on that Kubernetes layer, such as NVIDIA NIM™ microservices, NVIDIA TensorRT™ inference, NVIDIA Collective Communications Library (NCCL) collective operations, and multi-node training across NVIDIA NVLink-accelerated fabrics. PaletteAI provisions the clusters with NVIDIA GPU Operator and NVIDIA Network Operator built in, and helps keep the stack current as new versions ship.
“Buying GPUs is the easy part for an AI cloud operator. Operating them as a governed, multi-tenant service that stays profitable as you grow is the real work. This validation, with our other enterprise and government design validations, means operators can build that on a platform validated against NVIDIA’s requirements.” —Tenry Fu, CEO and Co-Founder, Spectro Cloud
Beyond Day 1
Validation confirms the platform aligns with the relevant NVIDIA reference architecture on the day you turn it on. Running one for the years that follow is the harder work: tenants stay live while hardware ages, and every new CVE needs remediation without downtime.
PaletteAI manages that lifecycle with reusable stack profiles and blueprint-driven automation. AI factories can take months to stand up; PaletteAI is designed to help operators move faster, and handles ongoing operations with the same automation.
PaletteAI is backed by technical support that helps coordinate support for common components such as NVIDIA AI Enterprise. For the rest of the ecosystem, we coordinate with the vendors and own the ticket from discovery to fix.
We already run large-scale AI factories for operators who cannot be named, including sovereign cloud environments in Europe and other highly regulated deployments. One regulated operator deployed an AI factory in a controlled environment on an approved timeline.
Talk to us about your AI factory
Whether you are standing up a neocloud, a sovereign AI service, or an internal AI factory, PaletteAI now carries NVIDIA AI Cloud Ready validation for AI infrastructure and platform software. Get started today.