TPU.sv is a reconstruction of Google's TPU (v1) in SystemVerilog. The primary objective of this project is to develop a generic and resource-adjustable machine learning inference accelerator for educational purposes.
The following sources have been primarily referenced for this project:
- Google's 2018 paper introducing TPU's architecture
- "Implementierung einer Tensor Processing Unit" by Jonas Fuhrmann
- tinyTPU by Jonas Fuhrmann
The core has been completed and successfully passes simulations as of 2024-10-12.
- Integration of an AXI module for communication
- Evaluation on real hardware with actual machine learning workloads
- Refinement of the hardware description
- Enhancement of the documentation
Contributions of any kind are welcome. If you encounter a bug, please do not hesitate to create an issue.
MIT License