Start tuning and serving AI models on a single DGX Spark with ready-made templates, then scale the same workflows to thousands of GPUs without a rebuild
RENO, Nev., June 30, 2026 /PRNewswire/ — CIQ, the enterprise software company behind Rocky Linux and the Fuzzball AI and HPC orchestration platform, today announced that its Fuzzball platform now delivers a production-ready AI compute and inference environment for the NVIDIA DGX Spark stack. AI teams get one consistent environment to develop, tune and deploy AI workloads on infrastructure they own and control. That environment runs the same way from a single Spark system to large GPU clusters and data-center infrastructure, with no rebuild as compute grows. NVIDIA DGX Spark is the first supported platform, with additional platforms planned.
The challenge teams now face is that before any model reaches production, AI teams must spend months assembling storage, container registries, schedulers, inference servers and deployment pipelines by hand. That assembly restarts every time the underlying compute changes, turning every infrastructure upgrade into a project that delays workload deployment and monetization.
Fuzzball replaces this with a ready-to-run AI compute and inference environment, designed specifically for private, local inference on DGX Spark. For organizations building sovereign AI workloads where data cannot leave the premises, that means a path from private model tuning to production inference that maintains control of the infrastructure underneath it. The time between a model idea and a running inference service compresses from months to days.
“Fuzzball is the Kubernetes of performance-intensive computing, and it is what AI teams have needed to truly own their infrastructure. The hard part of AI has never been the model. It has been operating that model at scale without rebuilding everything underneath it every time compute changes, and most teams spend months on that problem before a single workload reaches production. Fuzzball ends that. With hundreds of built-in workflow templates, a single DGX Spark becomes a complete environment for AI development, testing and validation from day one, and those same workflows run unchanged across thousands of systems and GPUs. CIQ stands for controlling intelligence. Fuzzball is how teams actually do it,” said Gregory Kurtzer, CEO and founder of CIQ and founder of Rocky Linux.
With Fuzzball, a single DGX Spark becomes a fully operational AI development and deployment environment from day one, with a direct path to larger infrastructure when the project demands it. Multiple DGX Spark systems can work together under a common Fuzzball environment, giving small teams a practical way to expand local compute capacity incrementally. When workloads grow beyond local capacity, the same containers, model assets and workflow definitions seamlessly move onto larger NVIDIA GPU deployments allowing you to scale from NVIDIA DGX Spark to NVIDIA GB300 NVL72 without changes to the application, Fuzzball orchestration model, or Fuzzball deployment process.
For AI teams and HPC teams that have historically managed separate toolchains, Fuzzball eliminates the operational split that slows AI from experimentation to production across the enterprise. Fuzzball bridges local DGX Spark systems and existing HPC clusters with cloud capacity and large GPU infrastructure including NVIDIA GPU systems, all under a single operational model. The result is a new generation of accelerated computing where job-based orchestration, workflow portability, containerized execution and production inference work together.
“Organizations in regulated industries have had to choose between moving AI to production and keeping data on infrastructure they control. That has never been an acceptable tradeoff. Fuzzball removes it. Teams can tune models privately on DGX Spark, operationalize those models as production inference services and expand onto larger controlled infrastructure without changing the environment underneath their work. For sovereign AI to be practical it has to run the same way at every compute tier, and that is exactly what Fuzzball on DGX Spark delivers. DGX Spark is the first platform this runs on, and it will not be the last,” said Bjorn Hovland, president of CIQ.
To see Fuzzball in action, join CIQ’s upcoming webinar, “How to deploy your own LLM and take it to production with Fuzzball” on Thursday, July 16, 2026 at 12pm PDT / 3pm EDT. Register at https://events.ciq.com/fuzzball-inferencing-ai-workload.
About CIQ
CIQ is the founding support and services partner for Rocky Linux and a leading provider of Enterprise Linux infrastructure. CIQ delivers commercially supported Linux offerings, high-performance computing solutions, and AI infrastructure to enterprises, government agencies, research institutions and supercomputing centers worldwide. CIQ’s products include the Rocky Linux from CIQ (RLC Pro) family of operating systems, Ascender Pro for IT automation, Fuzzball job-based container orchestration, Warewulf cluster provisioning and Apptainer, the leading container system for high-performance computing. For more information, visit ciq.com.
Media Contact:
Cristin Connelly
Cathey Communications for CIQ
cristin@cathey.co
View original content to download multimedia:https://www.prnewswire.com/news-releases/fuzzball-turns-nvidia-dgx-spark-into-a-ready-to-run-ai-development-and-inference-environment-at-any-scale-302814931.html
SOURCE CIQ
