Software
AMD Releases ROCm 7.13 with Expanded Hardware and Virtualization Support
AMD announced ROCm 7.13, supporting AMD Instinct MI350P and extending GPU virtualization on KVM and VMware ESXi.
Image: AMD
AMD has released ROCm 7.13, expanding hardware support across enterprise datacenters and introducing new tools for developers. The release includes ROCm Core 7.13, AMD GPU Driver 31.30, and AMD GPU Virtualization 9.0. These updates enable GPU virtualization on VMware ESXi and KVM, optimized performance for generative AI and large language models, and support for the latest AMD Instinct accelerators. The platform also introduces new profiling tools and a modular packaging model to streamline development workflows. According to AMD, the release aims to make AI development more practical across a broader range of hardware, from local prototyping to large-scale production environments. TheRock, AMD’s automated build and release system, continues to evolve, offering a base installation with optional expansion SDKs for specialized domains like HPC and data science. The new version also includes support for Ubuntu 24.04.4, Ubuntu 26.04, and RHEL 9.6 for the MI350P accelerator. *Source: [amd](https://rocm.blogs.amd.com/ecosystems-and-partners/rocm-7.13-blog/README.html)*
Key points
- AMD released ROCm Core 7.13, the AMD GPU Driver 31.30, and AMD GPU Virtualization 9.0.
- ROCm 7.13 expands hardware support across enterprise datacenters, introducing AMD’s latest Instinct accelerators.
- AMD GPU Virtualization 9.0 brings new capabilities to GPU partitioning support for KVM and newer ESXi versions.
- ROCm 7.13 includes new profiling and visualization capabilities, along with a modular packaging and installation model.
- AMD GPU Driver 31.30 enables the new MI350P GPU, expanding the hardware ecosystem.
- ROCm 7.13 validates MI350P on Ubuntu 24.04.4, Ubuntu 26.04, and RHEL 9.6.