The core of this architecture relies on the Cedar multimodal model and the Cedar Agent, both built upon Google’s Gemma. By utilizing the NVIDIA Jetson T5000, SDMC’s AI Station provides shared edge compute power, allowing multiple applications to run locally. This setup minimizes the need for high-performance hardware in every individual device while keeping latency-sensitive and private data within the home environment.
SDMC Integrates Google Cloud and NVIDIA Tech into Home AI Architecture
At IBC 2026 in Amsterdam, SDMC unveiled a deployable AI Home architecture that merges cloud-edge intelligence with Physical AI. Selected for Google Cloud’s Proactive Physical AI Acceleration Program, the company is bridging the gap between digital reasoning and real-world household actions through shared edge computing and multimodal models.
SDMC is also expanding its hardware ecosystem through partnerships with Amlogic, integrating Edge AI into displays, streaming boxes, and 4K projectors. Beyond entertainment, the company is applying these technologies to network management via the AI Fiber Gateway, which automates diagnostics and performance optimization. These developments aim to help operators scale services across existing home devices, moving beyond simple digital assistants toward autonomous home control and robotic interaction.
Comments (0)
No comments yet. Be the first!