Run secure LLMs locally with AMD Lemonade Server, powered by Ryzen™ AI, and XMPro’s Composable Intelligence Platform.
The new XMPro–AMD solution allows industrial organizations to eliminate cloud AI costs, reduce latency, and maintain full control over operational data by running powerful AI models directly at the edge using AMD’s Lemonade Server and XMPro’s composable intelligence platform.
Also Read: CIO Influence Interview with Dipto Chakravarty, Chief Product and Technology Officer at Black Duck
XMPro, a leading industrial intelligence platform provider, has announced a partnership with AMD, a global leader in high-performance computing, to deliver on-site AI processing directly to industrial facilities. The collaboration integrates AMD’s open-source Lemonade Server with XMPro’s composable platform, enabling manufacturers to run advanced AI models on-site without cloud dependencies, latency issues, or unpredictable costs.
This collaboration demonstrates the potential of AI at the industrial edge to drive measurable impact in mission-critical environments.”
— Pieter Van Schalkwyk – XMPro CEO
The solution addresses three critical challenges facing industrial AI adoption: unpredictable cloud costs, data sovereignty risks, and network dependency. A typical automotive manufacturing plant processing 50,000 sensor readings per minute can face over $15,000 monthly in cloud AI API costs alone¹. The new edge-based approach eliminates these recurring expenses while keeping sensitive operational data completely on-site.
“Our collaboration with XMPro is about more than just performance – it’s about enabling real-time intelligence where it matters most,” said Paul Hartke, Fellow, AMD Research and Advanced Development . “By combining AMD’s industrial edge hardware with XMPro’s composable AI platform, manufacturers can analyze and act on operational data instantly, without compromising security or data sovereignty. This is the next step in deploying practical, high-impact AI across the industrial landscape.”
The integrated solution combines AMD’s Lemonade Server, optimized for local large language model deployment, with XMPro’s composable industrial intelligence platform. While AMD provides the high-performance local AI processing power, XMPro delivers the critical industrial context and system integration that transforms raw AI capability into actionable business value.
XMPro’s platform enables companies to rapidly connect AI models to existing industrial systems—ERP, SCADA, IoT sensors—without disrupting current operations. Using drag-and-drop tools, engineering and operations teams can build data streams that unify information across their enterprise, create digital twins of assets and processes, and deploy AI-powered workflows that automatically detect anomalies, predict failures, and trigger intelligent responses.
Companies can now deploy comprehensive AI-powered decision support directly on their factory floor using AMD Ryzen™AI systems, which deliver up to 50 TOPS of AI compute performance through hybrid NPU and integrated GPU architecture.
Performance testing demonstrates significant advantages over traditional CPU-only setups. The AMD Ryzen™ AI hybrid configuration achieves up to 8-9 times faster token generation and 3-4 times faster initial response time across industry-relevant models². This performance improvement translates to faster AI decision-making on the factory floor, real-time responsiveness for AI-powered operational guidance, reduced hardware load and energy use during AI inferencing, and improved user experience for engineers and operators using AI assistants.
“By combining AMD’s industrial edge compute capabilities with XMPro’s composable AI platform, manufacturers can now run complex AI inference directly on-site, enabling faster decisions, greater operational autonomy, and complete control over sensitive data,” said Pieter Van Schalkwyk, CEO at XMPro. “This collaboration demonstrates the potential of AI at the industrial edge to drive measurable impact in mission-critical environments.”
The partnership delivers three key advantages that cloud-based AI cannot match:
• Complete Data Sovereignty: All AI processing occurs on-site, ensuring sensitive operational data never leaves the facility – critical for organizations with strict security requirements or regulatory compliance needs.
• Predictable Costs: One-time hardware investments replace recurring cloud API fees, providing cost certainty and eliminating budget surprises from variable usage.
• Zero Network Dependency: Local processing eliminates latency from cloud round trips and ensures AI capabilities remain operational even during network disruptions.
XMPro’s role in the partnership extends beyond platform capabilities to solve the practical challenge of operationalizing edge AI in industrial environments. While many companies can deploy local AI models, XMPro provides the industrial-specific framework to make those models immediately useful for real-world operations. The platform’s composable architecture allows organizations to create reusable AI workflows that can be rapidly deployed across multiple sites and adapted to changing business needs, turning edge AI from a technical capability into a scalable business solution.
The solution targets asset-intensive industries including manufacturing, energy, mining, utilities, and logistics, where real-time decision-making and data security are paramount. XMPro’s industrial-focused design includes pre-built connectors for common industrial systems, templates for typical use cases like predictive maintenance and process optimization, and governance frameworks that ensure AI decisions remain traceable and auditable, critical requirements for regulated industries.
AMD’s Lemonade Server is fully open-source and compatible with OpenAI APIs, ensuring easy integration with existing applications. The server runs efficiently on AMD Ryzen™ AI hardware, leveraging both the Neural Processing Unit (NPU) and integrated graphics processor (iGPU) for maximum performance and energy efficiency.
Also Read: Confidential Computing vs Traditional Encryption: Key Differences Explained
[To share your insights with us as part of editorial or sponsored content, please write to psen@itechseries.com]
