MulticoreWare, AMD Partner to Advance Physical AI and Robotics

Companies Demonstrated Real-Time Multimodal AI and Vision-Language-Action Workflows on AMD Ryzen AI Platforms at AMD Advancing AI 2026

 MulticoreWare, Inc., a world expertise firm specializing in AI software program options, bodily AI, accelerated computing, and engineering providers, as we speak introduced its ongoing collaboration with AMD to advance autonomous robotics and edge intelligence on AMD platforms.

As a part of this collaboration, MulticoreWare joined AMD at AMD Advancing AI 2026 to current ‘Enabling Physical AI on AMD’, demonstrating how superior imaginative and prescient, language, and motion (VLA) fashions can drive real-time robotic intelligence on AMD Ryzen AI Embedded platforms.

As AI more and more strikes from the cloud into robots, autonomous methods, and clever edge gadgets, organizations want environment friendly methods to run refined AI fashions nearer to the place selections want to be made. Together, AMD and MulticoreWare are serving to builders carry superior notion, reasoning, and motion capabilities to AMD-powered methods.

At AMD Advancing AI 2026, AMD and MulticoreWare demonstrated how multimodal VLA fashions run on AMD Ryzen AI Embedded built-in GPUs utilizing AMD ROCm, enabling robots to understand, purpose, and act in actual time. The session showcased sensible steerage for AI builders, robotics engineers, and innovators constructing next-generation clever machines on AMD Embedded platforms.

“Physical AI is reshaping how machines understand, resolve and act in the actual world,” stated Sumit Shah, Head of Product Management and Marketing, Adaptive and Embedded Computing Group, AMD. “AMD Ryzen AI Embedded X100 Series processors ship a scalable, open x86 Embedded platform that unifies AI, real-time management and industrial reliability to allow the era of autonomous methods with out locking builders right into a single compute structure or software program stack.”

“A Physical AI system is dependent upon a tightly built-in loop between notion and actuation. It should function in actual time, on actual {hardware}, and in environments which might be inherently unpredictable,” stated Vish Rajalingam, VP & GM, Mobility and Transportation BU at MulticoreWare. “That makes it a hardware-software co-design problem, not merely an AI inference drawback. Building on the open-source AMD Robotics Software Suite, we work carefully with OEMs to optimize all the stack in order that latency, reliability and accuracy targets are constantly achieved in manufacturing environments. That’s the combination MulticoreWare and AMD ship collectively to transfer clever robotic methods from prototype to deployment.”

This session builds on greater than 15 years of collaboration, with MulticoreWare delivering software program optimization, AI, and engineering experience throughout the AMD ecosystem, together with Ryzen AI, Ryzen, AMD EPYC, AMD Instinct, AMD Radeon, and adaptive computing applied sciences.

The publish MulticoreWare, AMD Partner to Advance Physical AI and Robotics first appeared on AI-Tech Park.