ENLIGHT™ deep learning accelerator

Supreme solution for AI edge computing

Precision/Area optimized AI accelerator technology

OPENEDGES provides proprietary precision/area optimized AI accelerator technology. Low area and low power consumption are vital for edge AI solution. OPENEDGES’ proprietary AI accelerator technology maps your neural network model to the AI accelerator processor “ENLIGHT” with very low area while preserving enough performance and precision so you can realize the best edge AI solution.

In order to realize AI based SoC which satisfies your requirement, just having the good AI accelerator performance is not enough but you need to achieve the total performance throughout the system, which includes on-chip communication and memory access. OPENEDGES provides all of the components for the AI based SoCs.

  • AI accelerator “ENLIGHT” which is a general purpose NPU with ultra-low area and power consumption,
  • NoC “OIC” which is a high performance and physical design friendly on-chip interconnect
  • Memory controller “OMC” which is a high performance and low area / power consumption

They are designed to work together to have optimal performance so you will be able to design the best performance AI based SoC with minimum risk.