Abstract
A number of two-dimensional LSIs (2D-LSIs) with the thickness of around 30 μm which play amplifying and converting image signal and some arithmetic operations are integrated vertically in order to achieve a real-time microvision system. This microvision system transfers the two-dimensional (2D) image data arrays as it is using high density vertical interconnections. So, the image information signals are processed in parallel in each LSI and the processings are performed in pipeline over all the system. In this study we design the test chip which plays edge detection by Laplacian operator. In CAD simulation, the processing time of the edge detection takes about 10 μsec using 2 μm CMOS design rule. In fabrication, grinding and chemical-mechanical polishing techniques are used to thin the wafer to 30 μm. The thinned wafer with buried interconnections is bonded vertically to a thick wafer through micro-bumps after careful alignment by the newly developed wafer aligner with the alignment tolerance of 1 μm. The microvision system with 3D integration structure can be fabricated by repeating such sequence.
| Original language | English |
|---|---|
| Pages | 831-835 |
| Number of pages | 5 |
| State | Published - 1996 |
| Event | Proceedings of the 1996 IEEE/SICE/RSJ International Conference on Multisensor Fusion and Integration for Intelligent Systems - Washington, DC, USA Duration: 1996.12.8 → 1996.12.11 |
Conference
| Conference | Proceedings of the 1996 IEEE/SICE/RSJ International Conference on Multisensor Fusion and Integration for Intelligent Systems |
|---|---|
| City | Washington, DC, USA |
| Period | 96.12.8 → 96.12.11 |
UN SDGs
This output contributes to the following UN Sustainable Development Goals (SDGs)
-
SDG 9 Industry, Innovation, and Infrastructure
Fingerprint
Dive into the research topics of 'Real-time microvision system with three-dimensional integration structure'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver