Iris Coleman October 13, 2024 02:37
AMD releases ROCm 6.2.3 to enhance AI capabilities on Radeon GPUs with enhanced support for Llama 3, Stable Diffusion, and Triton frameworks, improving AI development efficiency.
AMD has launched the latest version of its open computing software, AMD ROCm™ 6.2.3. It is specifically designed to enhance Radeon GPU performance on native Ubuntu® Linux® systems. According to AMD.com, this update aims to provide better inference performance for AI models, especially Llama 3 70BQ4, and will help developers bring Stable Diffusion (SD) 2.1’s text-to-image capabilities to their AI projects. It will be possible to integrate.
Main features of ROCm 6.2.3
The new ROCm 6.2.3 release offers several advanced features aimed at accelerating AI development.
Llama 3 support via vLLM: This feature provides superior inference performance on Radeon GPUs with Llama 3 70BQ4 models. Flash Attendant 2 integration: Designed to optimize memory usage and improve inference speed, this feature supports forward enablement. Support for Stable Diffusion 2.1: Developers can now incorporate SD text-to-image models into their AI applications. Triton Framework Beta Support: Enables developers to write high-performance AI code with minimal expertise and efficiently utilize AMD hardware.
Advances in AI development
Erik Hultgren, AMD’s software product manager, emphasized that ROCm 6.2.3 targets specific features to accelerate generative AI development. This release includes professional-level performance enhancements for Large-Scale Language Model (LLM) inference with vLLM and Flash Attendant 2. It also introduces beta support for the Triton framework, expanding the scope of AI development on AMD hardware.
Evolution of ROCm support
ROCm support for AMD’s Radeon GPUs has evolved significantly over the past year, starting with the 5.7 release. Version 6.0 expands functionality by incorporating the ONNX runtime and officially certifying more Radeon GPUs, including the Radeon PRO W7800. The 6.1 update marks a new milestone with support for multi-GPU configurations and integration with the TensorFlow framework.
In its current release, ROCm 6.2.3 continues to focus on Linux® systems and will soon introduce support for Windows® Subsystem for Linux® (WSL 2). This strategic approach aims to further strengthen the ROCm solution stack for Radeon GPUs and position it as a robust option for AI and machine learning development.
Visit AMD’s official community page for additional information and resources.
Image source: Shutterstock
Source link
