CIQ optimizes RLC Pro AI for DGX Spark with the newest long-term kernel and NVIDIA drivers, so the same Enterprise Linux runs from prototype to production.
RENO, Nev., October 15, 2026 - CIQ, the enterprise software company behind Rocky Linux, today announced RLC Pro AI (Rocky Linux from CIQ Pro AI) optimized for NVIDIA DGX Spark. Developers get leading AI performance from the newest long-term kernel and NVIDIA's latest validated GPU drivers, on the same trusted, vendor-backed Enterprise Linux that runs business-critical infrastructure. On Spark Arena, the community leaderboard where DGX Spark owners publish reproducible benchmark recipes, a DGX Spark running RLC Pro AI outperformed four common sparkrun tests at up to 10% faster than the leading entry.
The challenge developers face is that the operating system they choose for speed is rarely the one their organization runs in production. Development stacks keep up with the newest kernels and drivers, while Enterprise Linux has typically trailed them to protect stability. When a prototype is ready, teams move it to a different operating system and rebuild drivers, containers and dependencies, then check whether the performance survived the move. Every one of those moves delays the point where AI work starts to deliver value.
RLC Pro AI closes that gap with one operating system for both sides. It runs the CIQ Linux Kernel 6.18, the newest long-term kernel, with NVIDIA's latest validated GPU drivers to get more from the Spark's Grace Blackwell hardware. When that work moves to production, the speed moves with it, along with the lifecycle stability, security maintenance, and enterprise support production depends on.
AI performance starts with the operating system. The software beneath the AI stack decides how quickly a developer can test an idea, evaluate a model, and feed the results into the next experiment, which makes the choice of operating system a practical decision about AI productivity.
"Developers have had to pick between the fastest operating system and the one their company trusts in production. RLC Pro AI ends that choice, and provides an optimized, homogenous environment from prototype to production. It puts the newest long-term kernel and validated GPU drivers on Enterprise Linux, and it beat the Spark community's own leaderboard to prove it. Developers get the speed they want on their Spark today, and when that work becomes business-critical, it is already on the operating system enterprises run their most important infrastructure on," said Bjorn Hovland, President of CIQ.
Enterprise infrastructure matters from the first experiment. RLC Pro AI brings the lifecycle stability, security maintenance, and commercial accountability of Rocky Linux from CIQ into the development environment, so teams build on the same Enterprise Linux choice they'll rely on as applications grow.
RLC Pro AI also works with CIQ Fuzzball, which CIQ previously announced for DGX Spark, to take that work to larger GPU environments for training and inference.
Availability
RLC Pro AI optimized for NVIDIA DGX Spark is available now. Developers can register for a free RLC Pro AI developer license for their Spark through the CIQ Portal. The developer license is for individuals and covers up to 25 nodes for non-production, non-commercial use.