nearbAI on ASIC

nearbAI on ASIC

nearbAI – AI close to the sensors

The nearbAI semiconductor IP is a small, low-power and affordable AI core that runs locally, close to your sensors. This results in a low and predictable latency and runs with ultra low-power consumption.  easics’ embedded AI solutions integrate tightly with novel and existing sensors such as image sensors capturing light inside and outside the visible spectrum (such as hyperspectral and thermal infrared), 3D scanning laser (LiDAR), Time-of-Flight (ToF) sensors, radar, microscopy, ultrasound sensors, and microphones.

Why choose nearbAI as AI accelerator?

Low hardware cost with MAC efficiencies above 95%
Superb flexibility supporting multiple primitive CNN and RNN operations
Fast time to market to embed AI close to the sensor
100% Customizable to meet your performance, power consumption, area and memory requirements.

nearbAI software tools


Supported operations

The core is optimised by parameterisation of our generic core towards application specific needs. nearbAI supports the following operations:

  • convolution engine: 
    • 2D convolution
    • depth-wise convolutions
    • matrix multiplications
    • fully connected layers
    • dense layers
    • bias
  • configurable post-processor
    • Max pooling, average pooling
    • ReLU,RELU6, Leaky ReLu,
  • convolutional neural networks (CNNs), such as ResNet, YOLO, MobileNet
  • recurrent neural networks (RNNs), such as DeepSpeech

Which sensors benefit from nearbAI?

Image sensor: visual, near-IR, thermal IR, X-ray
Time-of-Flight (ToF), 3D stereo
LiDAR, radar
audio sensor
contact us to discuss your sensor application!
ASIC architecture SiP_7

Which customers benefit from nearbAI?

The nearbAI IP offers benefits to semiconductor companies and original equipment manufacturers (OEM).

Semiconductor companies
For semiconductor companies or sensor manufacturers we provide an AI solution for smarter sensors and structured data output. nearbAI can outperform AI on MCU for real-time decision making. Possible sensors include: image, audio, lidar and many more.
OEMs will benefit from nearbAI if it comes to outperforming classical vision algorithms and AI integration in their systems for cameras, vehicles, robotics, inspection machine and more. We offer an embedded solution for Deep Learning on ASIC, preferably as a System-on-Chip (SoC). Working with ASIC instead of FPGA, GPU or CPU offers lots of advantages in terms of performance, size and power.

nearbAI architecture

The DMA Controller loads input data and quantized weights into internal buffers. Both data and weights are then shifted through the Convolution Engine. The results are sent to the Accumulator and are finalized in the Post Processor. The Sequencer manages the execution of the subsequent layers of the DNN. It generates a continuous flow of tensors through all layers of the DNN. The final output tensors are returned as results —structured data— to your application.

nearbAI architecture and data flow

Deep learning on ASIC – download PDF documentation

Want to know more about nearbAI?

Download the product PDF here.