DT-MATCH: Defect- and Topology-Aware Chiplet Matching for Yield Optimization in Heterogeneous Integration

By Xuanyi Tan 1, Sanmitra Banerjee 1, Dhruv Thapar 1, Sreejit Chakravarty 2, and Krishnendu Chakrabarty  1
1 School of Electrical, Computer and Energy Engineering, Arizona State University, Tempe, AZ
2 Ampere Computing, Santa Clara, CA

Abstract

Heterogeneous integration (HI) and advanced packaging enable chiplet-based system-on-chip (SoC) designs; however, they also introduce new yield challenges arising from manufacturing defects and complex die-to-die (D2D) interconnect constraints. In particular, lane-level defects, limited spares, and strict structural compatibility requirements impede the efficient assembly of functional systems from defective but repairable chiplets. We present DT-MATCH, a defect- and topology-aware chiplet matching framework that maximizes the number of functional SoCs under realistic assembly constraints. DT-MATCH provides two complementary approaches. The integer linear programming (ILP)-based approach is implemented using a batching strategy to approximate the exact formulation. The greedy randomized adaptive search procedure (GRASP)-based approach provides substantial acceleration, but it does not guarantee yield optimality. Both approaches use a preprocessing pipeline to reduce the problem size. Experimental results on industrial designs demonstrate that, as expected, the ILP-based approach achieves superior solution quality. The GRASP-based approach leads to lower runtimes while still achieving solution quality that is comparable to the ILP-based solution. Together, these methods provide a trade-off between solution quality and efficiency for practical chiplet assembly optimization.

To read the full article, click here