DARPA-Backed Foundry TIE Aimsโ to pioneerโ 3D Chip Assembly by 2026
AUSTIN, TX – A new,โ first-of-its-kind fabricationโ facility (fab) focused on 3D Heterogeneous Integration (3DHI) is rapidly taking shape in Austin, Texas, with a goal of full operational โcapability in the first quarter of 2026. โคThe foundry, operatedโค by TIE (details of the full name not provided in source), is a key component of the Next Generation Manufacturing (NGMM) program funded by the Defense Advanced Research Projects Agencyโ (DARPA).
The facility, toured recently by IEEE spectrum, โis already seeing the installation of numerous chip manufacturing and testing tools, and has onboarded a growing team ofโค engineers and technicians. Unlike customary silicon foundries, TIE is โฃdesigned toโ be a “high-mix, low-volume” operation, โฃcapable ofโค handling a diverse range ofโ projects rather than mass-producing a singleโฃ chip type.
A meaningful challengeโ facing the โขfab is the inherent variability of non-siliconโข wafers – differences in size and thermal expansion rates – which must be overcome to achieveโข the โmicrometer-precision required for linking chips in 3D structures.To address โขthis, TIE is developing Process Design Kitsโข (PDKs) and Assembly Design Kits (ADKs). The ADK is โnotably crucial, providing the rulesโค for 3D assembly and advanced packaging techniques like hybrid bonding.
The initial manufacturing โคprocesses will be โฃrefined through three “exemplar” projects: โa phased-array radar, an infrared imager (focal plane array), and a compact power converter. “Piloting thoseโฃ through production givesโ us an initial roadmapโฆ an on-ramp into tremendous innovation โฃacross a broader request โspace,” explained a TIE executive, identified as Holmes in the article.
Because TIE won’tโค have the benefit of running large batches of similar test wafers to identify and correct process flaws – a common practice in high-volume silicon foundries – the company is leveraging artificial intelligence developed by Austin-based Sandbox Semiconductor toโฃ predict the impact of process adjustments.
The NGMM program is alsoโฃ fostering collaboration withโฃ academic institutions,โข including UT Dallas, whereโค researchers are focusing on areas like new thermal conductivity films, microfluidic cooling, and failure analysis in complexโฃ 3D packages. Dr. Ted Moise,an โIEEE Fellow and professor โat UT Dallas,described NGMM as “a very rare opportunity” for research.
DARPA’s Whitney Mason, director of the Microsystems Technology office, acknowledged the unusual nature โคofโค the program, stating, “It’s not our habit to stand up facilities that do manufacturing.” However, โขsheโค suggested the innovative spirit of Austin – โฃembodied in โthe city’s motto, โ”Keep Austin Weird” -โข makes it a fitting location for this pioneeringโฃ effort.