Event camera. Neuromorphic chip. One device.
The first edge AI module that processes event-camera data natively on neuromorphic silicon. No GPU. No frame conversion. Sensor to inference at under 1mW — with Nuro as the programming layer.
Early exploration — hardware partnerships in progress. Join the interest list to shape what we build.
The problem
The right tools exist. Nobody connected them.
Event cameras have no processing stack
Cameras like iniVation DAVIS or Prophesee EVK4 produce asynchronous spike streams. But every downstream system expects frames. Teams throw away the temporal structure by converting spikes back to images — then run a GPU model on them. That defeats the point.
Neuromorphic chips have no sensor story
Intel Loihi 2 and SpiNNaker 2 process spikes natively, but ship as bare chips with no sensor integration. Users wire event cameras to neuromorphic boards manually, with custom glue code, every time.
The software layer doesn't exist
No SDK takes event-camera output and runs it on neuromorphic hardware end-to-end. Researchers maintain two or three codebases and spend weeks on integration instead of their actual work.
Why hybrid vision
Pure event cameras are powerful but hard to use.
Pure event cameras
- ·Microsecond temporal resolution
- ·High dynamic range (120dB+)
- ·Sparse output — low bandwidth
- ·No spatial context in static scenes
- ·Hard to integrate with existing pipelines
Hybrid vision sensors
- ·Events + frames from the same pixel array
- ·Spatially aligned — no registration needed
- ·Frames give context, events give timing
- ·SNN processes events; frames supervise
- ·Works with existing computer-vision tooling
The full stack
Photon to inference, in four layers.
Hybrid vision sensor
Frames + events
A next-gen sensor outputs a standard image frame and an asynchronous event stream simultaneously — pixel-aligned, from the same silicon. Spatial context from frames, microsecond timing from events. No information lost.
Nuro SDK
Dual-stream SNN
Nuro ingests both streams natively. Events feed directly into a spiking network — no conversion. The frame channel provides spatial context. Train the dual-stream model on GPU with surrogate gradients.
Neuromorphic processor
Loihi 2 / SpiNNaker 2
The trained network runs on a neuromorphic chip co-located with the sensor. Neurons compute only on a spike — no clock waste on silence. Always-on inference at under 1mW. Chip and sensor speak the same language.
Vantar Cloud
Develop + monitor
Develop and benchmark remotely before deploying to the physical module. Monitor energy draw, spike rates, and latency from anywhere. Over-the-air model updates when you retrain.
Use cases
Always-on vision at the edge, without a GPU.
Robotics — always-on perception
Drones and mobile robots need collision avoidance that never sleeps and never drains the battery. Event cameras detect motion at 1μs; neuromorphic inference runs on milliwatts. The Dev Kit is the perception module.
Industrial inspection
High-speed production lines move too fast for frame cameras. Event cameras capture micro-defects at microsecond resolution; neuromorphic processing gives real-time classification without a GPU rack at every station.
AR/VR — low-latency tracking
Head and hand tracking needs sub-millisecond latency and can't afford GPU inference on a battery-powered headset. Events + neuromorphic cut latency ~10× and power ~100× vs frame-based tracking.
Automotive — edge ADAS
Event cameras handle high dynamic range and fast motion better than frame cameras. Running inference on neuromorphic silicon removes power-hungry edge GPUs from the perception stack.
Why not a GPU
The right tool for training. The wrong tool for the edge.
Help us build
the right thing.
The Dev Kit is in early exploration. We're talking to robotics engineers, drone teams, and edge AI researchers about the form factor and API they need. If it's relevant to your work, join the list — we'll reach out directly.