Generative AI’s prominent application is AI image generation, with Stable Diffusion models leading this domain. However, running these models locally has been complex. To simplify this, AMD partnered with TensorStack to optimize AI image generation on AMD hardware. Their Amuse 2.0.0 beta program leverages AMD Ryzen AI 300-series processors and Radeon RX 7000-series graphics, requiring high-performance systems with substantial RAM. Key features include AMD XDNA Super Resolution for upscaling images and an AI Filters drawer for creating custom filters, all with a user-friendly interface.
Generative AI’s most popular application is AI image generation. Stable Diffusion models are the backbone of local image generation and pioneered this field. However, existing solutions for running these models on AI PCs are often cumbersome, requiring technical expertise and optimizing performance.
To address this, AMD has collaborated with four-year old TensorStack, an AI technology firm based in New Zealand, to offer what AMD says is a seamless, user-friendly, and optimized experience for customers with AMD Ryzen AI 300-series processors and Radeon RX 7000-series graphics. This partnership streamlines the process, making AI image generation more accessible and efficient.
AMD’s recently released Amuse 2.0.0 beta program is designed to allow users to generate high-quality images using AI models on modern AMD hardware-based PCs. To harness the full potential of Amuse, users need a premium system equipped with either AMD Ryzen AI 300-series processors and a minimum of 24GB of RAM, or a Radeon RX 7000 system boasting at least 32GB of memory. Additionally, systems built around Ryzen 8040-series processors are also supported, provided they have at least 32GB of RAM. This underscores that Amuse is intended for modern, high-performance systems, delivering exceptional image generation capabilities.
AMD saysAmuse 2.0.0 beta unlocks a world of creative possibilities. It allows users to produce stunning images with AI models, digitize paintings and drawings, and craft custom AI filters.
One standout feature of AMD Amuse is its utilization of AMD XDNA Super Resolution, which upscales images from 512×512 to 1024×1024, streamlining the image generation process. With Amuse, says AMD, users can bring their imagination to life by creating original images, transforming hand-drawn art into digital masterpieces and applying or designing custom AI filters to enhance existing images.
Amuse also has an AI Filters drawer, where users can create their own AI filters and change the look and feel of existing images using generative AI. This is all done locally using AMD Ryzen AI and Radeon graphics.
According to AMD, Amuse 2.0.0 is an easy-to-install program that has no dependencies and a simplified user interface, called Ez Mode, that will take care of setting up the configuration required for AI image generation. Some key features are:
- No command line needed
- Simple and easy to use
- AMD XDNA Super Resolution
- Single executable
- No dependencies
- Autoconfiguration and model selection
- Paint to Image
- Drawing to Image
- Create your own AI filters
- Multi-stage, ONNX-based pipeline (Stable Diffusion, Control Nets, Feature Extractors)
Amuse 2.0 is a beta release and experimental software. It is being provided on an as-is basis. As such, there may be bugs, and the user experience may be unstable or cause unintended behavior. Furthermore, there is a potential risk that the models may produce content that heavily draws from copyrighted material, sparking concerns about copyright infringement and unauthorized sharing. Moreover, the absence of an Internet connection to identify and verify the sources of training data may also pose a problem, potentially leading to unintended copyright violations.
Creativity will know no bounds in the AI PC era.
