Analyzing Various Structural and Temperature Characteristics of Floating Gate Field Effect Transistors Applicable to...

Area of Science:

• Semiconductor Device Physics
• Computer Architecture
• Materials Science

Background:

• Traditional von Neumann architecture faces limitations in data processing and energy efficiency.
• Logic-in-memory (LiM) computing aims to overcome these limitations by integrating logic and memory functions.
• Floating gate field-effect transistors (FGFETs) are explored as a promising device for LiM applications due to their compatibility with existing silicon processes.

Purpose of the Study:

• To analyze the temperature-dependent characteristics of FGFET source drain barrier (SDB) and central shallow barrier (CSB) devices.
• To confirm the applicability of these FGFETs for LiM computing.
• To evaluate the performance of LiM circuits designed with FGFETs, including ternary content-addressable memory (TCAM) and full adder (FA) circuits.

Main Methods:

• Device and circuit simulations using TCAD and HSPICE at the 32 nm technology node.
• Development of compact models for FGFET-SDB and -CSB devices.
• Design and analysis of TCAM and FA circuits utilizing FGFETs.
• Investigation of FGFET device potential for neural network applications.

Main Results:

• FGFET-SDB and -CSB devices exhibit temperature-dependent characteristics suitable for LiM applications.
• CMOS-compatible FGFETs offer an advantage over novel materials for LiM computing.
• TCAM and FA circuits designed with FGFETs were analyzed for energy and delay time performance.
• The study identified the optimal number of CSBs for specific applications.

Conclusions:

• FGFETs are a viable technology for developing efficient LiM computing systems.
• The temperature-dependent analysis provides crucial insights for designing robust LiM circuits.
• Further exploration of FGFETs for neural network applications is warranted.