AMD has introduced a broad range of AI-focused products during its Advancing AI 2026 event, expanding its portfolio across data centers, enterprise, edge, and robotics. Leading the announcement is AMD Helios, the company’s first rack-scale AI solution, alongside new EPYC server processors, Instinct GPUs, embedded AI processors, and robotics platforms. AMD says the new lineup is designed to address the growing demand for agentic AI and large-scale inference workloads while maintaining an open ecosystem that gives customers greater flexibility.
AMD Helios Rack Scale AI Solution
The centerpiece of AMD’s announcement is the Helios rack-scale AI solution, which combines 72 AMD Instinct MI455X GPUs with 18 sixth-generation EPYC “Venice” processors and AMD Pensando networking. According to AMD, the platform delivers up to 30% more inference tokens per dollar than competing rack-scale AI solutions, making it suitable for large-scale AI deployments. Several technology companies, including OpenAI, Anthropic, Meta, Microsoft, Oracle, and Supermicro, are already working with AMD to deploy or support Helios-based infrastructure.
AMD 6th Gen EYPC GPU and Instinct MI400 Series GPUs
AMD also introduced its sixth-generation EPYC processors, which are built to handle cloud computing, enterprise applications, high-performance computing, and AI workloads. Complementing the CPUs are the new Instinct MI400 Series GPUs, including the MI455X for AI inference and training, the MI430X for scientific computing and sovereign AI, and the MI350P, which is designed to deliver AI acceleration in existing infrastructure. Together, these products form the foundation of AMD’s next-generation AI compute platform while supporting the company’s roadmap through 2030.
ROCm.ai and partner ecosystem expand AMD’s AI platform
To strengthen its software ecosystem, AMD introduced ROCm.ai, a development platform that uses AI-assisted tools to simplify GPU programming and optimize workloads across AMD hardware. The platform is designed to work with coding assistants such as Claude, Codex, and Cursor while supporting popular AI frameworks including PyTorch, Hugging Face, vLLM, and SGLang. AMD also highlighted collaborations with companies such as Anthropic, OpenAI, Meta, AT&T, Cisco, and Cerebras to accelerate AI infrastructure, enterprise deployments, and software optimization.
AMD expands into physical AI and robotics
Beyond the data center, AMD announced new hardware aimed at physical AI and robotics. The company unveiled the Ryzen AI Embedded X100 Series processors together with new Kria AI System-on-Modules and the Kria AI Robotics Developer Platform, which combines CPU, GPU, NPU, and FPGA compute into a single development platform. AMD says these additions are intended to help developers accelerate robotics development while extending AI capabilities from cloud infrastructure to intelligent machines operating in the physical world.
Grant is a Financial Management graduate from UST. His passion for gadgets and tech crossed him over in the industry where he could apply his knowledge as an enthusiast and in-depth analytic skills as a Finance Major. His passion allows him to earn at the same time help Gadget Pilipinas' readers in making smart, value-based decisions and purchases with his reviews and guides.
