About cookies on this site Our websites require some cookies to function properly (required). In addition, other cookies may be used with your consent to analyze site usage, improve the user experience and for advertising. For more information, please review your options. By visiting our website, you agree to our processing of information as described in IBM’sprivacy statement. To provide a smooth navigation, your cookie preferences will be shared across the IBM web domains listed here.
Publication
DPC 2024
Conference paper
Reliable Direct Bonded Heterogeneous Integration ( DBHI) for AI Hardware
Abstract
With a continuous focus to maximize computational performance by optimizing package size and reducing fabrication costs, this paper presents the most recent advances on the technology of DBHi (Direct Bonded Heterogeneous Integration) packages as a chiplet packaging technology. This is one of the promising solutions for AI hardware as it can provide high bandwidth and low power communication between chiplets. In this study, we present DBHi with very fine chiplet interconnection pitch of 30 and 75µm pitch with two different DBHi architectures called Standard DBHi and Surface DBHi. These two DBHi configurations are tested for thermomechanical stresses during deep thermal cycling (DTC). The finite element model calculated with these architectures confirms overall robustness of Surface DBHi configuration. In this work, we present the effects and the advantages that came with Surface DBHi i.e. thinning down the Si bridge and the removal of cavity in the laminate for bonding process and their influence on the overall thermomechanical performance of the package.