BufFormer: A Generative ML Framework for Scalable Buffering

Buffering is a prevalent interconnect optimization technique to help timing closure and is often performed after placement. A common buffering approach is to construct a Steiner tree and then buffers are inserted on the tree based on Ginneken-Lillis style algorithm. Such an approach is difficult to scale with large nets. Our work attempts to solve this problem with a generative machine-learning (ML) approach without Steiner tree construction. Our approach can extract and reuse knowledge from high quality samples and therefore has significantly improved scalability. A generative ML framework, BufFormer, is proposed to construct abstract tree topol-ogy while simultaneously determining buffer sizes & locations. A baseline method, FLUTE-based Steiner tree construction followed by Ginneken-Lillis style buffer insertion, is implemented to generate training samples. After training, BufFormer can produce solutions for unseen nets highly comparable to baseline results with a correlation coefficient 0.977 in terms of buffer area and 0.934 for driver-sink delays. On average, BufFormer-generated tree achieves similar de-lays with slightly larger buffer area. And up to 160X speedup can be achieved for large nets when running on a GPU over the baseline on a single CPU thread.