Reevaluating Google's Reinforcement Learning for IC Macro Placement

Does anyone know how this is related to the recently discussed work on AlphaChip [1]?

[1] https://news.ycombinator.com/item?id=41672110