CIQ and AMD Optimise Rocky Linux for AI and HPC Workloads

CIQ and AMD are to optimise enterprise infrastructure for AI and HPC workloads running on AMD datacentre solutions. It will start with an AMD optimised version of Rocky Linux from CIQ with validated AMD drivers and ROCm support. The first phase will also include AMD Instinct GPUs and the AMD ROCm software platform. The next phase will feature additional AMD optimisations throughout CIQ’s infrastructure stack.

Gregory Kurtzer, CEO of CIQ and Founder of Rocky Linux, said, “Enterprise customers expect to move from infrastructure deployment to workload execution quickly. This collaboration gives AMD a single, reproducible Linux foundation to optimize against, and it gives enterprises a path to deploy AMD datacenter solutions from day-zero, without procurement hurdles.

“Rocky Linux is already the OS of choice for performance-intensive computing. Adding AMD-specific optimization and keeping it freely accessible makes that combination even stronger for AI and HPC workloads.”

Instinct GPUs driving Enterprise AI and HPC

Buying separate hardware and software stacks might have worked for general-purpose computing, but AI demands tighter integration. AMD is keen to position its hardware, along with a complete software stack for enterprise AI. Partnering with CIQ gives AMD a wider appeal to customers and a complete offering that they need.

CIQ brings its own commercially supported Rocky Linux (RLC) operating system. It comes in two versions, RLC Pro and RLC Pro AI, which was announced two weeks ago. It contains a pre-validated NVIDIA stack and can be deployed in just under 4 minutes.

AMD is seeing a significant increase in the adoption of its AMD Instinct GPUs. They are being adopted for AI training, inference and HPC. It has signed several deals for its use in AI infrastructure with other vendors.

In June last year, AMD announced that Oracle was to deploy Instinct MI355X GPUs in OCI. Microsoft also said its deployment of the Instinct MI300X50 was powering AI models on Azure.

In October, OpenAI announced it was to deploy 6 gigawatts of Instinct MI450 Series GPUs starting in June 2026. And just last month, Meta announced that it was also going to deploy 6 gigawatts of Instinct GPUs for the next generation of its AI infrastructure.

CIQ’s RLC Pro AI will now support AMD Instinct GPUs in the same way it supports NVIDIA. For customers, that makes this deal attractive. It gives them the confidence that they are able to deploy at speed, at scale and with a deeper integration between software stack and hardware.

Adding to that is the standardisation of a validated foundation for AMD ROCm software. It will simplify deployment and lifecycle management. At the cluster scale for HPC, image management and version alignment can slow deployments. Validated, reproducible OS builds reduce these operational bottlenecks.

Why CIQ and Rocky Linux?

Rocky Linux has become one of the most widely deployed enterprise Linux distributions in the world. There are currently tens of millions of Rocky Linux systems in use globally. That gives end-users access to a wider trained workforce, which reduces the cost of deployment and maintenance. It also delivers more apps that will run seamlessly on the OS.

In the last three months, CIQ has made nine separate announcements around Rocky Linux. Most are around new releases and wider product support. In addition to RLC Pro AI, it has also released a major cryptography update.

The CIQ Network Security Services (NSS) module for Rocky Linux has achieved the Cryptographic Algorithm Validation Certification (CAVP). It is now on the Modules in Process (MIP) list. It also delivers two NIST-approved Post-Quantum Cryptography (PQC) algorithms. This allows customers to use Rocky Linux in their most sensitive environments.

Importantly, it is also FIPS 140-3 compliant. That means it will enable government buyers to put AMD and Rocky Linux on their infrastructure list.

Enterprise Times: What does this mean

Enterprise customers want highly integrated platforms on which they can build solutions. Those platforms are increasingly a combination of hardware and software. The better the integration, the faster the time to ROI. In a world where ROI is becoming more elusive, offering an option that can improve that, will always appeal to customers.

But there is often a trade-off. The tighter the integration, the more risk there is of it becoming a proprietary lock-in. By working with CIQ, AMD is ensuring that it avoids that challenge. The tighter integration this offers with RLC Pro AI also brings additional benefits.

It means that customers who have developed on NVIDIA-based architectures can now move across to AMD Instinct architectures. That is because of the pre-validated stacks that RLC Pro AI has for both vendors’ GPU architectures.

It will be interesting to see how soon the two companies announce customers for this new configuration. And, importantly, who those customers are. We may hear more in a month’s time at the AMD AI DevDay in San Francisco.

The post CIQ and AMD Optimise Rocky Linux for AI and HPC Workloads appeared first on Enterprise Times.