AMD Brings Microsoft Project Zenith to Ryzen AI Halo for Powerful Local AI Development on Windows 11
AMD has confirmed that Microsoft’s Project Zenith is now available on its Ryzen AI Halo developer platform, giving developers a ready-to-use Windows 11 environment built for software creation, AI experimentation, and local model development.
The platform is powered by AMD’s Ryzen AI Max+ PRO 495 and is available in a compact mini PC design with up to 192GB of unified memory. That makes it one of the most memory-rich local AI development systems currently aimed at developers who want to work with large models without relying entirely on cloud services or expensive discrete GPU workstations.
Project Zenith is not a new version of Windows. Instead, it is a preconfigured Windows 11 development setup designed to reduce the time developers usually spend installing tools, adjusting settings, and preparing a machine for coding or AI work.
Out of the box, Project Zenith includes a broad selection of commonly used developer tools. These include Visual Studio Code, Git, GitHub CLI, GitHub Copilot CLI, Azure CLI, Python, Node.js, .NET, PowerShell 7, and Windows Subsystem for Linux. Developers can still customize the system after setup, but the goal is to provide a faster starting point for building apps, testing AI models, and managing development workflows.
AMD is also pairing the system with its own AI software resources. The Ryzen AI Developer Center gives users access to AMD-validated tools, documentation, and configurations, while AMD AI Playbooks provide guided workflows for different AI tasks.
These playbooks cover several popular use cases, including running large language models with PyTorch and ROCm, generating images through ComfyUI, using Qwen3-Coder for local coding assistance, deploying LLMs through LM Studio, and setting up automation workflows with n8n.
The biggest attraction of the new Ryzen AI Halo platform is its memory capacity. The system uses AMD’s Ryzen AI Max+ PRO 495, a high-end chip featuring 16 Zen 5 CPU cores, 32 threads, and boost clocks of up to 5.2 GHz. It also includes Radeon 8065S integrated graphics with 40 RDNA 3.5 compute units, offering up to 61 FP16 TFLOPS of performance.
For dedicated AI acceleration, the chip includes an XDNA 2 NPU rated at 55 TOPS. However, the standout feature for local AI workloads is the 192GB of LPDDR5X unified memory. AMD says up to 160GB of that memory can be allocated to the GPU, supported by 273 GB/s of memory bandwidth.
This large unified memory pool is important because many advanced AI models require enormous amounts of memory, especially when working with larger parameter counts, extended context windows, or multiple workloads running at the same time.
AMD claims the 192GB Ryzen AI Halo configuration can support AI models with up to 300 billion parameters. In practice, this would require heavy quantization and the right software configuration, but the memory capacity gives developers far more room to experiment compared with typical consumer systems.
The new model represents a step up from the earlier Ryzen AI Halo configuration, which used the Ryzen AI Max+ 395 and offered up to 128GB of unified memory. The increase to 192GB should be particularly useful for AI developers, researchers, and advanced users who want to test larger language models locally.
That said, this is clearly a premium developer platform rather than a mainstream desktop replacement. A 192GB Linux-equipped Ryzen AI Halo system has been listed at $7,299.99, while the original 128GB version has been listed at $3,999.99. AMD says Project Zenith is available preinstalled on the new hardware, though availability of a 192GB Windows retail configuration may vary by retailer and region.
Existing Ryzen AI Halo owners using Windows are not left behind. AMD says they can recreate the Project Zenith-style development environment by using Microsoft’s open-source Windows Developer Config project.
With Project Zenith on Ryzen AI Halo, AMD and Microsoft are trying to make local AI development more practical and accessible. The combination of Windows 11, preinstalled developer tools, AMD’s AI software ecosystem, and a massive unified memory pool gives developers a powerful alternative to cloud-based AI infrastructure.
For those building AI apps, testing large language models, experimenting with image generation, or creating local coding assistants, the Ryzen AI Halo platform offers a serious all-in-one development environment designed for the next generation of AI workloads.
