Edgecortix
About Us
EdgeCortix’s Dynamic Neural Accelerator (DNA) architecture and industry first AI hardware and software co-exploration engine enable adaptation of neural inference co-processor designs in an application specific manner, bringing cloud-level performance to resource constrained edge devices, for ultra-low latency with batch size=1, low cost and energy-efficient deep neural network inference. We offer our proprietary DNA neural inference co-processor IP, co-exploration software and compiler, designed to deliver from 6 TOPs to greater than 100 TOPs with a small power budget as full SoC, while avoiding ultra-low bit-width quantization. DNA is designed to be hyper-scalable with flexible programming starting from high-level frameworks like Tensorflow and Pytorch.
Connect with us if you are interested in partnering with us on applications, evaluating our IP & software on FPGA prototypes or simply learning more about what is possible with DNA.
Chat with us during Hotchips:
We will be available online to talk with you live via our conference Slack channel (#s-edgecortix) between 8:00 AM and 6:00 PM (Pacific), Monday and Tuesday, August 17th and 18th.
- If you require a different time or to setup a private discussion with your team, please contact us.