AMD has released AMUSE 3.0, an update to its AI software that brings speed optimizations for Radeon RX 9070 graphics cards and Ryzen AI processors. The update aims to improve photo and video generation capabilities to meet the growing demand for advanced AI models. AMUSE 3.0 was developed in collaboration with TensorStack and supports a wider range of hardware.
It also introduces new AMD-optimized AI models from Stability.AI that promise speed improvements of 1.4x to 3.3x. The update is optimized for RDNA3.5 and RDNA4 architectures and includes features such as improved inpainting with SDXL support, smarter memory management, infinite upscaling via tiling support, Locomotion Text-to-Video support, and smarter model downloads with queuing and resume capabilities. To use AMUSE 3.0, users need specific hardware and software configurations, including AMD Radeon Graphics Cards and AMD Ryzen AI Processors.
They also need specific drivers from AMD and AI models from Stability.AI. The setup process requires some preparation, such as verifying hardware compatibility before installation. AMD has shared performance slides showing the benefits of AMUSE 3.0 on Radeon and Ryzen hardware.
Users can expect enhanced photo editing, video restyling, and even short video generation. AMUSE 3.0 also introduces new functionalities such as SDXL ControlNet capabilities and advanced Text-to-Video support. The AMUSE 3.0 software and the accompanying AI models are available for download.
Users should make sure their systems meet the necessary specifications to get the most out of the update. Stability AI has also worked with AMD to deliver select ONNX-optimized versions of the Stable Diffusion model family. These models are designed to run faster and more efficiently on Ryzen AI APUs.
AMD-optimized versions of Stable Diffusion 3.5 Large, Stable Diffusion 3.5 Large Turbo, Stable Diffusion XL 1.0, and Stable Diffusion XL Turbo are now available on Hugging Face and have the suffix “_amdgpu”.
Amuse 3.0 enhances Radeon RX 9070
The AMD-optimized versions of Stable Diffusion 3.5 deliver up to 2.6x faster inference compared to the base PyTorch models.
The Stable Diffusion XL models achieve up to 3.8x faster inference with AMD optimization. Amuse, the free AI art collaboration with TensorStack, has been updated with new features, including video generation. The update also includes new, optimized models specifically for AMD Ryzen processors.
Amuse provides AMD users with robust AI art generation tools, but it’s not exclusive to AMD hardware. Users with Intel Core PCs equipped with Nvidia discrete GPUs can also try out the range of Amuse’s capabilities. The newly added features include photo filters, local video creation, and AI-powered video generation and restyling.
These updates make Amuse a compelling alternative to cloud-based services like Midjourney, Microsoft Copilot, and Google Gemini. Amuse works best with AMD hardware, although it’s not a strict requirement. It’s optimized for standalone AMD discrete GPUs like the Radeon 9700 series and Ryzen AI 300 notebooks.
According to AMD, optimized models can generate images up to 4.3 times faster on the Radeon 9700 XT and up to 3.3 times faster on Ryzen AI processors at FP16 precision. The application offers two modes: “EZ Mode” for beginners and “Expert Mode” for advanced users. EZ Mode provides basic functionalities, while Expert Mode unlocks more advanced features and allows users to download additional AMD-optimized models for enhanced performance.
During testing, a prompt using the Ryzen-optimized model required just 270 seconds to generate, compared to 685 seconds with the standard model. Although the default resolution for text-to-video capabilities is 512×512, with more keyframes for longer videos, pushing the limits may affect the coherence of the generated art. Amuse offers a sophisticated range of functionalities in an easy-to-use interface, especially for AMD users.
While it may not be perfect, its capabilities make it one of the better AI art app suites available, not just for AMD users but for anyone interested in exploring the potential of AI-generated art.
Image Credits: Photo by Vishnu Mohanan on Unsplash
Rashan is a seasoned technology journalist and visionary leader serving as the Editor-in-Chief of DevX.com, a leading online publication focused on software development, programming languages, and emerging technologies. With his deep expertise in the tech industry and her passion for empowering developers, Rashan has transformed DevX.com into a vibrant hub of knowledge and innovation. Reach out to Rashan at [email protected]
