668704 event 1691008095 1691008137 <![CDATA[Ph.D. Dissertation Defense - Ankit Kaul]]> Title:  Modeling and Optimization of 2.5-D and 3-D Integration Architectures for Compute-In-Memory Applications

Committee:

Dr. Muhannad Bakir, ECE, Chair, Advisor

Dr. , Co-Advisor

Dr. Arijit Raychowdhury, ECE

Dr. Azad Naeemi, ECE

Dr. Suman Datta, ECE

Dr. Vanessa Smet, ME

Dr. Hemanth Dhavaleswarapu, AMD

]]> The objective of this research is to investigate power delivery network (PDN) and thermal management constraints in 3-D heterogeneous integration (HI) architectures for compute-in-memory (CIM) applications. First, we investigate design trade-offs in the power delivery network (PDN) of bridge-chip based 2.5-D heterogeneous platforms. We demonstrate that including a PDN in the bridge-chip can provide significant reduction in DC-IR drop, Ldi/dt noise, and high-frequency ripple compared to the baseline. Second, a comprehensive design-space exploration of PDN design for 3-D-HI CIM hardware. A methodology is proposed to evaluate and quantify trade-offs between power delivery design parameters and CIM performance metrics. Next, a device-integration methodology is proposed to quantify the thermal-driven impact of integration architectures on RRAM reliability for CIM applications. Subsequently, we present the thermal evaluation of a back-end-of-line (BEOL)-embedded chiplet integration scheme (polylithic 3-D), where custom chiplets are embedded into the BEOL of a primary tier. Finally, we present a study to evaluate the power delivery constraints for polylithic 3-D integration of BEOL-embedded chiplets. The potential impact of this research and potential future directions are summarized.

]]> <![CDATA[]]> https://teams.microsoft.com/l/meetup-join/19%3ameeting_MGFmYzM3MjYtMTJlYi00N2JjLTk1OTctYzA2OTJkMGM2ZjU5%40thread.v2/0?context=%7b%22Tid%22%3a%22482198bb-ae7b-4b25-8b7a-6d7f32faa083%22%2c%22Oid%22%3a%222ba0bcb3-4c69-44e5-b0cc-ac5baa6cac1e%22%7d 434381 1788 100811 1808