CIQ expands Fuzzball to span five clouds & on-prem

CIQ has expanded its Fuzzball orchestration platform to support full multi-cloud deployments across CoreWeave, AWS, Google Cloud, Oracle Cloud and Microsoft Azure.

The platform also runs on on-premises infrastructure through the same control plane.

The move extends Fuzzball beyond single-environment orchestration for artificial intelligence and high-performance computing workloads. Customers can define a workflow once for training, inference or HPC jobs, then run it across different cloud providers or on their own systems without changing the workflow definition.

At runtime, the platform evaluates available environments and routes each job to the most suitable destination based on cost, performance and data locality. This is intended to reduce the operational work involved when organisations add new cloud environments to existing AI and HPC setups.

A genomics team that validates a sequencing pipeline on AWS can move it to Azure or Oracle Cloud without altering the workflow. A training job that needs Nvidia H100 capacity could be directed to CoreWeave, while a data-sensitive simulation could remain on-premises under policy controls.

Single control plane

At the centre of the launch is a provider-agnostic workflow definition that describes compute jobs, data movement, container images and resource requirements without cloud-specific logic. Fuzzball then translates that definition into infrastructure for the chosen environment, whether a hyperscale cloud, CoreWeave or a local cluster.

The system federates across five cloud environments as well as on-premises clusters. That gives users access to external GPU capacity and internal infrastructure through one control plane rather than separate deployment processes for each provider.

CIQ also outlined a common security model across the different environments. Each cloud deployment is provisioned through a two-phase automated process and uses one identity and access management model, one set of role-based access control policies and one secrets management posture across all supported clouds.

The platform removes static credentials through native identity services on each platform, including Workload Identity on Google Cloud, Managed Identities on Azure, Dynamic Groups on Oracle Cloud and IAM Roles on AWS.

Gregory Kurtzer, Chief Executive Officer and Founder of CIQ, said the product was designed to address the complexity facing AI teams working across different infrastructure environments.

"AI teams today are asked to ship faster, control costs and maintain sovereignty over their data, simultaneously, across infrastructure that was never designed to work together," said Kurtzer. "We built Fuzzball to solve that problem at the architectural level. When your workflow definition abstracts its requirements properly, you get portable access to every GPU environment the market offers and the freedom to route to wherever the best price, performance and data policy lives. Controlling your infrastructure and workloads is what enterprise AI infrastructure requires for production, and no other platform delivers it."

Reducing complexity

Multi-cloud strategies have become more common among AI developers and HPC operators as they seek access to scarce GPU supply, lower costs and greater control over where data is processed. But running workloads across multiple clouds often means rebuilding pipelines, rewriting scripts and repeating validation work for each provider.

CIQ is positioning Fuzzball as a way to remove that duplication by keeping workflow definitions, container images, data orchestration and job sequencing consistent across environments. That would let teams move workloads between providers without separate toolchains or provider-specific operational models.

Bjorn Hovland, President of CIQ, said the operational burden of multi-cloud had been one of the main barriers for many organisations.

"Fuzzball turns multi-cloud from a liability into a competitive advantage," said Hovland. "Five clouds used to mean five IAM models, five deployment pipelines and five sets of operational overhead, with complexity and risk multiplied."

The launch also reflects the growing importance of orchestration software in AI infrastructure, particularly as demand for graphics processing resources pushes companies to spread workloads across specialist cloud providers, major hyperscalers and in-house systems. CoreWeave has emerged as a significant supplier of GPU-rich cloud capacity, while large cloud providers continue to compete for AI and HPC workloads with broad infrastructure portfolios.

CIQ is best known as a support and services partner for Rocky Linux, the open-source enterprise Linux distribution. Its wider portfolio includes Linux operating systems, IT automation tools, cluster provisioning software and Apptainer, a container system used in high-performance computing.

Fuzzball's multi-cloud deployments for CoreWeave, AWS, Google Cloud, Oracle Cloud and Azure are available now. On-premises deployment remains supported on Warewulf, VMware and bare-metal clusters.