Despite decades of research in design automation,
circuit board design remains slow, tedious, and error prone.
~50%
of a typical electrical engineer’s time is spent on layout
$46B
estimated annual global spend on designing circuit boards
47 years
since the first commercial autorouter was released, still without significant industry adoption today
Quilter uses reinforcement learning informed by physics simulations to create a fully automated, superhuman circuit board designer.
Quilter designs circuit boards 100x faster than humans and delivers new iterations in hours, not weeks.
Quilter deeply optimizes board designs to eliminate extra design margins that add manufacturing costs.
Quilter uses physics simulations to validate real-world performance and eliminate delays from simple mistakes.
“At Quilter, we’re developing some of the most exciting industry applications of reinforcement learning. It’s a great place to be a researcher.”
~50% of Quilter’s total headcount is invested in exploring and implementing novel approaches and solutions to automated placement and routing.
“I joined Quilter because I saw how serious they were about designing a solution that embraces the messy physics that EEs consider when they design boards today.”
Quilter’s physics simulator is a comprehensive, full-wave Maxwell solver, used to identify and prevent issues and errors in PCB designs.
