Journal
IEEE TRANSACTIONS ON ELECTRON DEVICES
Volume 70, Issue 3, Pages 1029-1033Publisher
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TED.2023.3239330
Keywords
Iron; FinFETs; Performance evaluation; Films; Logic gates; Electric potential; Capacitors; Ambipolar; content addressable storage; fin field-effect transistor (FinFET); metallic source and drain (MSD); non-volatile memory
Ask authors/readers for more resources
An ultra-dense one-transistor (1T) ternary content addressable memory (TCAM) array is reported, based on high-performance, non-volatile, and ambipolar ferroelectric (Fe) silicon-on-insulator (SOI) fin field-effect transistors (FinFETs). TCAM functions were realized on a single device due to the multistates in the Fe layer and ambipolar characteristics. This has potential for Hamming-distance computing in area-and energy-efficient artificial intelligence (AI) processors.
An ultra-dense one-transistor (1T) ternary content addressable memory (TCAM) array is reported that is based on high-performance, non-volatile, and ambipolar ferroelectric (Fe) silicon-on-insulator (SOI) fin field-effect transistors (FinFETs). Because of the multistates in the Fe layer and the ambipolar characteristics, TCAM functions were realized on a single device. From the advanced CMOS process and the metallic source and drain (MSD) engineering, a maximum driving ON-current of similar to 0.1 mu A/mu m and a similar to 1000 ON/OFF ratio at V-GS = -0.25 V were realized for the 1T TCAM cell. This indicated large-array integration feasibility. In addition, 2 x 2 and 1 x 8 TCAM arrays are demonstrated. The quasi-linear dependence of a matching line (ML) voltage with the number of mismatched bits indicated a potential for Hamming-distance computing in area-and energy-efficient artificial intelligence (AI) processors.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available