Edge Computing Semiconductors Startups

DeepSilicon develops software and hardware tools that enable neural networks to run with improved speed and reduced computational costs.

4 investors

Stellon Labs develops compact AI models optimized for edge computing and resource-constrained hardware environments.

4 investors

Synapse Semiconductor is a semiconductor research lab building image sensors that run neural-network computation directly on the sensor, integrating light capture and AI processing in each pixel to replace GPU-based edge vision. Its chips target computer vision applications in self-driving cars, drones, humanoid robotics, and space systems.

3 investors

Kailash Labs builds compact, task-specific video reasoning models by distilling frontier vision-language models and post-training small open models for individual customer objectives, including on-premise deployment. Its first product, Marlin, is a 2B-parameter open-weights video VLM aimed at tasks such as dense captioning and temporal grounding for organizations processing large volumes of video.

1 investor

Tensil develops software that automatically designs custom hardware accelerators optimized for specific machine learning models, aiming to improve performance per watt-dollar compared to GPUs and other accelerators. The company targets product builders who want to embed efficient ML processing into their devices.

1 investor