What reinforcement learning brings to the table and where it can be used.
Node scaling and rising complexity are increasing the time it takes to get chips out the door. At the same time, design teams are not getting larger. What is needed is a way to automate the creative process, and to not have to start every design from scratch. This is where reinforcement learning fits in, with its ability to centralize and store “tribal knowledge. Thomas Andersen, vice president for AI and machine learning at Synopsys, talks about the need to learn the behavior of designs, and how machine learning can be used for everything from optimizing ATPG patterns and analog circuits to improved manufacturing.
 |
 |
 |
 |
 |
 |
 |
 |
 |  |
Leave a Reply