Synchron, NVIDIA Reveal Cognitive AI Brain Foundation Model for BCI

Synchron, a brain-computer interface (BCI) company, unveiled its roadmap to Chiral, a foundation model for human cognitions.

This marks the rise of Cognitive AI—artificial intelligence that’s trained directly on human neural activity. The company plans to advance BCI from supervised to self-supervised learning by combining large-scale neural data with advanced NVIDIA AI-powered computing.

Synchron showed at GTC 2025 that its AI-enabled BCI powered by the NVIDIA Holoscan platform could be displayed on the Apple Vision Pro. This meant people could control digital and physical environments using Synchron’s Stentrode direct thought-control technology.

One highlight was Synchron’s release of footage showcasing its BCI tech’s real-world impact. The video featured Rodney, a Synchron BCI user with ALS, operating and controlling his digital home environment entirely hands-free, voice-free, and touch-free. Check it out below:

The company’s collaboration with NVIDIA centers on three major tech milestones on its roadmap toward Cognitive AI.

1. Motor inference powered by NVIDIA Holoscan

Step one is to achieve real-time neural decoding with minimal latency. NVIDIA Holoscan augments on-device edge computation for faster, more precise motor inference in the BCI.

This lets users control digital environments with only thought, building a foundation for more advanced, high-dimensional interactions and AI enablement.

2. Context-aware interaction with NVIDIA Cosmos

The next step is integrating environmental awareness. NVIDIA’s Omniverse platform and Cosmos world foundation models generate physics-bound, photorealistic simulations of domestic environments to create richly labeled datasets that boost motor interference accuracy, adaptability, and fine-tuning.

Interpreting neural signals in real-world simulations allows the system to have higher-dimensional inference. This makes interactions more predictive, adaptive, and intuitive.

3. From neural data to Cognitive AI: Training Chiral

Synchron and NVIDIA will then train Chiral based on deidentified data. As a pre-trained brain model, Chiral will be engineered to abstract human cognitions from large-scale neural data.

As more devices are deployed, the model continuously improves, allowing higher-dimensional intent translation and evolving into a self-improving, general-purpose Cognitive AI, the companies believe. The transition surpasses intent recognition, laying a foundation for real-time intention-to-action.

“We are building a brain foundation model using generative pre-training techniques that learn directly from neural data—abstracting human cognition at its source—to create features that improve our user’s lives,” said Tom Oxley, MD, Ph.D., founder and CEO of Synchron. “This is possible because of our ability to scale large datasets, by making BCI as common as a stent insertion.”

“We are witnessing a new computing paradigm, where technology is no longer just responding to people—it’s empowering them in entirely new ways,” added David Niewolny, Senior Director of Business Development at NVIDIA. “By combining its breakthrough BCI technology with NVIDIA AI and Omniverse, Synchron is helping to deliver new possibilities for individuals with disabilities, enabling greater independence, communication, and connection with the world.”