Projects

Articles

Interests

Machine Learning

Design Flow

FPGA SoC Prototyping design flows

Technology

Accelerators

Authored Comments

Subject	Comment	Link to Comment
Almost.. 😀	Hi John, Thanks for your comment. Yes! I have designed an IP on CNN parallel acceleration and first verified it on FPGA. To learn GDS synthesising using SKY130A PDK and open-source tools, I have used that CNN IP with OpenLane RTL2GDSII flow and generated a physical design layout with all preliminary checks cleared. However, as of today, the project is still far away from sign-off for tapeout, as the design needs to be tested extensively for its characteristics. There is a steep learning curve for using open-source tools for design characterization, and I am actively working on it. Regards!	view

User statistics

My contributions

My comments

Comments

Tape Out ?

Hi,

Did I see that you designed a custom-built Convolutional Neural Network (CNN) Accelerator Engine IP as a part of your postgraduate studies and you taped it out on 130-nm process node using the Skywater Technology SKY130A PDK and open-source tools?

Almost.. 😀

Hi John, Thanks for your comment.

Yes! I have designed an IP on CNN parallel acceleration and first verified it on FPGA. To learn GDS synthesising using SKY130A PDK and open-source tools, I have used that CNN IP with OpenLane RTL2GDSII flow and generated a physical design layout with all preliminary checks cleared.

However, as of today, the project is still far away from sign-off for tapeout, as the design needs to be tested extensively for its characteristics. There is a steep learning curve for using open-source tools for design characterization, and I am actively working on it.

Regards!

Add new comment

To post a comment on this article, please log in to your account. New users can create an account.