A new technical paper titled “PPAC Driven Multi-die and Multi-technology Floorplanning” has been published by Texas A&M University and Duke University. The study focuses on floorplanning in heterogeneous integration, where different dies utilize distinct technologies. The paper highlights the importance of simultaneous technology selection in multi-die floorplanning, considering factors such as performance, power, and cost in addition to area and wirelength. The researchers developed a simulated annealing method and reinforcement learning techniques, which significantly outperformed a naive baseline approach in experimental results. This systematic study provides valuable insights into multi-die and multi-technology floorplanning, with potential implications for improving efficiency and performance in integrated circuits. Interested readers can access the full technical paper titled “PPAC Driven Multi-die and Multi-technology Floorplanning” by Roman-Vicharra, Cristhian, Yiran Chen, and Jiang Hu on arXiv, with the publication date of February 2025.
Source
Note: The image is for illustrative purposes only and is not the original image of the presented article.