At IBC 2026, SDMC is showcasing AI Station, a shared in-home AI compute platform for multiple AI models, devices, and services. As part of SDMC’s broader AI Home strategy and continued engagement across the Google ecosystem, AI Station helps operators extend AI capabilities across existing devices and services.

By allowing multiple AI workloads to share one platform, AI Station reduces the need for dedicated high-performance AI hardware in every endpoint. Latency- and privacy-sensitive workloads can stay at the home edge, while cloud intelligence extends capabilities when needed.

High-Performance Shared AI Compute

Powered by NVIDIA Jetson T5000, AI Station delivers up to 2,070 TFLOPS of Sparse FP4 AI performance with 128GB LPDDR5X memory, providing the headroom for demanding AI workloads and multi-model processing on one platform.

Beyond raw compute, AI Station provides a shared environment for private knowledge retrieval, multimodal content processing, and agent-driven execution—from private RAG and semantic retrieval across photos, videos, documents, and camera history to AI-driven actions across connected devices and services.


Enabling AI-Powered Home Services

Built on this compute foundation, SDMC AI Home solutions extend existing device ecosystems with higher-value applications across AI Home Security, AI Family Memory, and AI Operation.

Camera services can move beyond basic capture and motion detection to video understanding, event search, summaries, and intelligent alerts. Family photos and videos can become easier to organize, search, and rediscover through semantic understanding. AI-assisted operation can support more efficient device management, diagnostics, and execution across connected home environments.

For operators, this provides a practical way to extend AI capabilities across existing home ecosystems while introducing new AI-enabled service layers.


Designed for Flexible Deployment

SDMC combines system design, hardware-software integration, AI model adaptation, industrial design, and end-to-end project delivery to adapt AI Station to different deployment requirements.

This allows AI Station to serve as a common technology platform while supporting customer-specific implementations across different products and service environments.


Exploring Multimodal Intelligence for AI Home

At IBC 2026, SDMC is also demonstrating a technology preview of the Cedar multimodal AI model, built on Google’s Gemma and running on AI Station.

Designed as a compact, resource-efficient edge model, Cedar brings voice, text, vision, device states, and environmental context into a shared understanding of the home. The model is being developed around multimodal context understanding, contextual reasoning, and agent-based orchestration, with Cedar Agent supporting coordinated execution across devices and services.

For tasks requiring broader reasoning or cloud-scale services, Gemini complements edge intelligence in the cloud, forming a flexible cloud-edge model architecture for AI Home.


Scaling AI Home from Infrastructure to Services

By connecting shared in-home compute, cloud-edge intelligence, and existing device ecosystems, AI Station gives operators a foundation for expanding AI Home services, while offering brand customers a configurable path to differentiated AI products and solutions.

Looking ahead, SDMC will continue advancing AI computing infrastructure and multimodal intelligence to move AI Home toward a broader, service-driven ecosystem.

Meet us at IBC 2026, Booth 1.B33 to experience AI Station powering AI Home scenarios and explore how it can support your next AI Home deployment.