The AMD Radeon RX 7900 XT now supports ROCm 5.7
The AMD Radeon RX 7900 XT now supports ROCm 5.7, enabling PyTorch ML and AI acceleration.
- Why Enterprise RAID Rebuilding Succeeds Where Consumer Arrays Fail?
- Linus Torvalds Rejects MMC Subsystem Updates for Linux 7.0: “Complete Garbage”
- The Man Who Maintained Sudo for 30 Years Now Struggles to Fund the Work That Powers Millions of Servers
- How Close Are Quantum Computers to Breaking RSA-2048?
- Why Windows 10 Users Are Flocking to Zorin OS 18 Instead of Linux Mint?
- How to Prevent Ransomware Infection Risks?
- What is the best alternative to Microsoft Office?
The AMD Radeon RX 7900 XT now supports ROCm 5.7, enabling PyTorch ML and AI acceleration.
Last month, AMD released the ROCm 5.7.1 driver update for Ubuntu Linux, bringing ROCm and PyTorch support to the Radeon Pro W7900 and Radeon RX 7900 XTX graphics cards.
Prior to the AMD Advancing AI event in early December, AMD announced that ROCm and PyTorch support is now available for the RX 7900 XT graphics card, which shares the RDNA 3 architecture with the aforementioned products, covering all top-tier products of the RDNA 3 architecture.
According to AMD Software Product Manager Erik Hultgren, “We are excited about the latest additions to our product portfolio. These high-end GPUs, combined with ROCm, make artificial intelligence more accessible from both a software and hardware perspective, allowing developers to choose solutions that best suit their needs.”
Similar to the other two graphics cards, the RX 7900 XT features a massive 20GB of VRAM and a total of 168 AI acceleration units on the Navi 31 chip, providing fast machine learning training and inference performance. The GPU also supports PyTorch machine learning models and algorithms, optimized for AMD and compatible with Ubuntu Linux systems running version 22.04.3.
This is hopefully just the beginning, as AMD has many other RDNA 3 architecture GPUs. Consumers are hoping for ROCm support, as their competitors, NVIDIA and Intel, have consistently provided better general computing support on consumer-grade graphics cards, such as CUDA and the open-source Intel Compute Runtime. In comparison, AMD’s ROCm support is currently quite limited.
About ROCm 5.7.
ROCm is an open-source software platform developed by AMD to enable high-performance computing (HPC) and accelerate the development of GPU computing.
It is designed to provide a common platform for accelerating computing tasks across AMD GPUs. ROCm includes a set of programming interfaces, tools, and libraries for GPU computing.
Key features and components of ROCm typically include:
-
Heterogeneous Computing Support: ROCm is designed to support both multi-core CPUs and AMD GPUs, allowing developers to create applications that leverage the computational power of both types of processors.
-
HSA (Heterogeneous System Architecture): ROCm is built on the HSA standard, which aims to create a uniform platform for both CPU and GPU computing. This allows for better integration and cooperation between the CPU and GPU.
-
HIP (Heterogeneous-compute Interface for Portability): HIP is a ROCm component that provides a set of tools and conversion utilities to help developers port CUDA-based applications to ROCm. This eases the process for developers familiar with CUDA to transition their code to ROCm.
-
OpenCL™ and Vulkan® Support: ROCm supports industry-standard APIs such as OpenCL and Vulkan, providing a versatile environment for GPU computing and graphics development.
-
Deep Learning Frameworks: ROCm includes support for popular deep learning frameworks such as TensorFlow and PyTorch. This allows developers to perform accelerated machine learning and AI tasks on AMD GPUs.
-
Version Updates: ROCm receives periodic updates to enhance performance, add support for new hardware, and improve compatibility with the latest software frameworks.
Please note that for the most accurate and up-to-date information, it’s recommended to check the official AMD ROCm documentation or AMD’s official announcements. You can visit the ROCm GitHub repository for the latest releases and details.
