Accelerate Silicon Design with Jupyter Notebooks

Explore how to accelerate silicon design using Jupyter Notebooks in this 41-minute tutorial by Johan Euphrosine from Google. Learn to leverage Conda packages for publishing and sharing interactive silicon design experiments. Discover how to run design experiments from conception to GDS using publicly-hosted notebooks without local tool installation. Gain insights into scaling notebooks on public cloud providers to explore parameter spaces of various silicon designs. Follow the process of deploying an open-source Terraform solution for provisioning Jupyter notebooks with pre-installed tools. Understand how to report estimated performance metrics to a blackbox and hyperparameter optimization service for suggesting new parameters. Observe the rapid convergence of experiments towards optimal metrics for given designs. Explore the creation of standalone notebooks for sharing, aggregating, and reproducing experiments. Dive into topics such as hardware synthesis, Open Lane configuration, PIN floor planning, routing, signoff, and multiplication. Gain practical knowledge on configuration options, visualization techniques, and the OpenMPW shutter process.

Intro
Who are you
Mission at Google
Vision
Open Design Elements
Software Chain
Packages
Sharing Notebooks
Demo
Hardware Synthesis
Hardware Synthesis Example
Open Lane
Configuration
PIN
Floor Plan
Routing
Signoff
Multiplication
Open Link
Configuration Options
Preview
Visualization
Aggregate
OpenMPW Shutter