Thanks to visit codestin.com
Credit goes to github.com

Skip to content

Open-source AI Accelerator Stack integrating compute, memory, and software — from RTL to PyTorch.

License

Purdue-SoCET/atalla

Repository files navigation

This is "Atalla 0x01", a student-led initiative to design and implement a 
complete **AI Hardware-Software** stack -- from RTL to PyTorch integration. 

This project aims to become an open, reproducible, and 
research-grade infrastructure that unites all key accelerator 
components under one roof.

Current synthesis flows are built on MITLL90nm, but will soon 
be upgraded to TSMC65nm. 

The main source tree includes these subdirectories:
    /docs/ -> Documentation of Arch, Contribution Guide, etc. 
    /kernels/ -> Benchmarking kernels.
    /rtl/ -> Core Design Files. 
    /tb/ -> Formal, Functional and UVM verification. 
    /waves/ -> .do scripts for Questa.
    /scripts/ -> CI/CD pipelines
    /reports/ -> Genus Optimized Reports.
    /Flowkit/ -> SoCET-private Physical Design Flows with Genus/Innovus 
    /aihw-ppci-compiler/ -> Compiler Infra.
    /atalla-sim/ -> Cycle-Accurate Simulator.
    /UVM_1.2/ -> UVM 1.2 Core Libs.
    /LICENSE 
    /Makefile  

If you have questions, please send mail to [email protected].

Enjoy using Atalla 0x01 and please share your modifications and extensions.

About

Open-source AI Accelerator Stack integrating compute, memory, and software — from RTL to PyTorch.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Contributors 14