NVIDIA CUDA 13.2 Released with Expanded Support
NVIDIA has officially announced support for Red Hat Enterprise Linux (RHEL) compatible distributions, including AlmaLinux, with the release of CUDA 13.2. This move allows users to receive direct support from NVIDIA when utilizing their infrastructure on these platforms.
Meanwhile, AlmaLinux has confirmed this development, stating that they now have an agreement with NVIDIA to distribute NVIDIA packages directly from their repositories. Therefore, users will no longer need to obtain drivers separately, ensuring a more streamlined experience.
Benefits of Official Support
The official support from NVIDIA for RHEL-compatible distributions like AlmaLinux means that users can expect more synchronized updates. For example, when NVIDIA releases new drivers, AlmaLinux can now update their packages in tandem, eliminating brief version mismatches.
Additionally, this agreement enables AlmaLinux to ship open-source drivers alongside userspace and CUDA components, ensuring that all packages are updated simultaneously. However, it’s essential to note that this change may take some time to be fully implemented.
Furthermore, CUDA 13.2 introduces several new features, including a spin-wait dispatch mode to reduce execution latency, new PTX features, and improved C++20 standards conformance with the NVCC compiler. These updates aim to enhance the overall performance and usability of CUDA.
Conclusion and Next Steps
In conclusion, NVIDIA’s decision to support RHEL-compatible distributions with CUDA 13.2 is a significant step forward for users of these platforms. As a result, users can expect a more integrated and supported experience when using NVIDIA infrastructure on AlmaLinux and other compatible distributions.
For more information on this development, users can visit the AlmaLinux blog for detailed insights and updates on the latest CUDA release.
