Red Hat integrates Nvidia CUDA into Enterprise Linux and OpenShift
Red Hat will henceforth distribute Nvidia's CUDA Toolkit directly via its platforms. This should simplify the provision of GPU-accelerated AI applications.
(Image: heise medien)
Red Hat and Nvidia have announced an expanded partnership that integrates the CUDA Toolkit directly into Red Hat's product portfolio. Developers will be able to access the essential tools for GPU-accelerated applications via the official repositories of Red Hat Enterprise Linux (RHEL), OpenShift, and Red Hat AI. This is intended to simplify installation and automatically resolve dependencies.
The integration addresses a central challenge in the use of AI systems in enterprise environments: the operational complexity in the interplay of different components. Developers have so far had to invest considerable effort to identify compatible drivers, manage dependencies, and reliably run workloads on different systems. Direct distribution via Red Hat's platforms largely eliminates this integration effort.
The CUDA Toolkit includes compilers, libraries, and developer tools for programming on Nvidia GPUs. It is now provided via a unified, tested software stack that, according to Red Hat, offers a consistent environment for AI workloads, regardless of whether they are run locally, in public clouds, or at the edge. This approach aligns with Red Hat's existing hybrid cloud strategy.
Videos by heise
Red Hat emphasizes the open source nature of the collaboration. The company positions itself as a bridge between the open hybrid cloud and Nvidia's AI platform without building a closed ecosystem. The integration is intended to allow companies to retain freedom of choice in tools and technologies while simultaneously providing a stable and secure platform.
Red Hat did not provide details on the availability and exact timeline of the integration in the announcement.
(fo)
