GitPedia

Free TPU

Free TPU for FPGA with compiler supporting Pytorch/Caffe/Darknet/NCNN. An AI processor for using Xilinx FPGA to solve image classification, detection, and segmentation problem.

From embedeep·Updated May 30, 2026·View on GitHub·

After four years R&D, EEP-TPU has evolved into the second-generation architecture with V3+ version, and has been embedded in three ASIC chips to achieve mass production. The project is written primarily in Shell, distributed under the MIT License license, first published in 2018. Key topics include: caffe, cnn-accelerator, darknet, deep-learning, fpga.

After four years R&D, EEP-TPU has evolved into the second-generation architecture with V3+ version, and has been embedded in three ASIC chips to achieve mass production.

FREE TPU V3plus for FPGA has updated, which is the free version of a commercial AI processor (EEP-TPU) for Deep Learning EDGE Inference.

User can download lastest EEP-TPU from https://github.com/embedeep/FREE-TPU-V3plus-for-FPGA

Contributors

Showing top 3 contributors by commit count.

View all contributors on GitHub →

This article is auto-generated from embedeep/Free-TPU via the GitHub API.Last fetched: 6/24/2026