Synchron and NVIDIA Collaborate on Real-Time Neural Processing for BCIs
Synchron, a developer of implantable brain-computer interface (BCI) technology, has announced a collaboration with NVIDIA to enhance its next-generation BCI using the NVIDIA Holoscan platform. The partnership, announced at the 2025 JP Morgan Healthcare Conference, aims to improve real-time neural processing and explore large-scale brain-language modeling.
According to Synchron, the company plans to leverage NVIDIA Holoscan in two main areas:
- It will enhance real-time edge AI processing by improving signal processing and multi-AI inference capabilities directly on the implantable device. This is expected to reduce latency, strengthen data privacy, and create a more responsive BCI experience. NVIDIA Holoscan reportedly provides a unified framework supporting various AI models and data modalities, optimized for seamless sensor integration and accelerated real-time processing.
- Synchron states that it intends to develop a large-scale foundation model for brain inference by processing neural data at scale. The goal is to create scalable, interpretable brain-language models that could improve neuroprosthetics, cognitive expression, and user interaction with digital devices.
In July 2024, Synchron integrated generative AI into its brain implant system to improve communication for individuals with severe disabilities, including those with amyotrophic lateral sclerosis (ALS). The update uses OpenAI’s multimodal generative pre-training transformer (GPT) to enable text and audio generation from brain signals. The system generates automated prompts for texting and chatting based on contextual inputs, including emotional states, aiming to match the speed of normal conversation. Synchron has stated that user-generated inputs are not shared or stored with OpenAI, maintaining privacy and data security.
Synchron has been working on implantable BCI technology since 2019 and has reportedly completed two human clinical trials. The company is now preparing for a larger-scale clinical study with a device optimized for scalable manufacturing. Synchron’s technology is designed to restore motor function and expand autonomy for individuals with paralysis. The company also reports having an exclusive partnership with flexible MEMs manufacturer Aquandas to support its device development.
The company also released a detailed overview of its roadmap for Chiral, a foundation model that abstracts human cognition from large-scale neural data. The outline explains a phased approach: beginning with real-time neural decoding, progressing to context-aware interactions, and ultimately training the model on deidentified clinical data. This framework reflects a measured transition from supervised to self-supervised learning, forming a neutral basis for further integration of neural signals into cognitive AI systems.