AMD is excited to release one-step diffusion models that demonstrate the readiness of AMD Instinct™ MI250 accelerators for model training and further research. The models are designed to provide performance comparable to traditional full-step diffusion models while maintaining the efficiency needed for training on data center systems or deployment on edge devices, such as AI-enabled PCs and laptops.

 Full technical blog AMD Nitro Diffusion: One-Step Text-to-Image Generation Models

The AMD Alveo™ V80 compute accelerator brings massive hardware parallelism and memory bandwidth for workloads with large datasets. Visit us at SC24 to learn how this solution can help you accomplish more with HPC and AI. 

ROCm 6.3 marks a significant milestone for the AMD open-source platform, introducing advanced tools and optimizations to elevate AI, ML, and HPC workloads on AMD Instinct GPU accelerators. ROCm 6.3 is engineered to empower a wide range of customers—from innovative AI startups to HPC-driven industries—by enhancing developer productivity

This blog delves into the standout features of this release, including seamless SGLang integration for accelerated AI inferencing, a re-engineered FlashAttention-2 for optimized AI training and inference, the introduction of multi-node Fast Fourier Transform (FFT) to revolutionize HPC workflows and more. Explore these exciting updates and more as ROCm 6.3 continues to drive innovation across industries.

Amuse 2.2 Beta is now available for download. 

A step-by-step guide on how to run Stable Diffusion 3.5 Large, Stable Diffusion 3.5 Large Turbo or Stable Diffusion 3.5 Medium on AMD Radeon™ Graphics Cards and Ryzen™ AI 300 Series Processors. 

In recent years, the rapid development of artificial intelligence technology, especially the progress in large language models (LLMs), has garnered significant attention and discussion. From the emergence of ChatGPT to subsequent models like GPT-4 and Llama, these language models have demonstrated remarkable capabilities in natural language processing, generation, understanding and reasoning. Continuing AMD tradition of open-sourcing models and code to help the community advance together, we are excited to release our first series of fully open 1 billion parameter language models, AMD OLMo.

To dive deeper into the three stages of training, and AMD OLMo model results, please reference the full article here: Introducing the First AMD 1B Language Models: AMD OLMo

See how AMD Ryzen™ AI 300 series-based PCs compare to the competition in llama.cpp based applications like LM Studio. 

AMD has just released the latest version of its open compute software, AMD ROCm™ 6.2.3 which supports Radeon GPUs on native Ubuntu® Linux® systems. 

Fine-tuning a large language model (LLM) is the process of increasing a model's performance for a specific task. This blog will introduce you methods and benefits on fine-tuning Llama model on AMD Radeon GPUs.

In the ever-evolving landscape of artificial intelligence, large language models (LLMs) like GPT-4 and Llama have garnered significant attention for their impressive capabilities in natural language processing and generation. However, small language models (SLMs) are emerging as an essential counterpart in the AI model community offering a unique advantage for specific use cases.  AMD is excited to release its very first small language model, AMD-135M with Speculative Decoding.  This work demonstrates the commitment to an open approach to AI which will lead to more inclusive, ethical, and innovative technological progress, helping ensure that its benefits are more widely shared, and its challenges more collaboratively addressed.