A 6-Gbps 16-nm FinFET CMOS I/O Buffer With Variation Insensitivity Ensured by Genetic Algorithm

This paper presents a novel 6-Gbps variation insensitive input/output (I/O) buffer designed for DDR4 and DDR5 SDRAM data transfer in a 16-nm FinFET CMOS process. Utilizing genetic algorithm (GA) to model process, voltage, and temperature (PVT) variations, the study reveals insights into temperature...

Full description

Saved in:
Bibliographic Details
Published inIEEE transactions on circuits and systems. I, Regular papers Vol. 71; no. 11; pp. 4961 - 4972
Main Authors Wang, Chua-Chin, Chodisetti, L S S Pavan Kumar, Ke, Jhih-Ying, Lo, Cheng-Yao, Lee, Tzung-Je, Tolentino, Lean Karlo Santos
Format Journal Article
LanguageEnglish
Published New York IEEE 01.11.2024
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Subjects
Online AccessGet full text

Cover

Loading…
More Information
Summary:This paper presents a novel 6-Gbps variation insensitive input/output (I/O) buffer designed for DDR4 and DDR5 SDRAM data transfer in a 16-nm FinFET CMOS process. Utilizing genetic algorithm (GA) to model process, voltage, and temperature (PVT) variations, the study reveals insights into temperature and voltage effects on FinFET-based, nanoscale buffer characteristics, leading to the removal of the temperature detector circuit to save power and area. Voltage variations, however, significantly impact slew rate, prompting the introduction of a Voltage Detector circuit using ultra-low threshold voltage (ULVT) transistors. Innovative Voltage Level Converter, Pre-Driver, and Digital Logic Control circuits enhance slew rate and throughput while stabilizing the output signal quality. This results in reliable operation at 6.0 Gbps with improved slew rate (17.7%/39.75% for VDDIO =0.8/1.2 V) and duty cycle performance (50.5%/51.4% for VDDIO =0.8/1.2 V) due to PV auto-adjustment; the first in the world. The proposed design effectively addresses the stringent slew rate and data rate requirements of DDR4 and DDR5 SDRAMs, offering advancements in speed, reliability, and efficiency amidst PV variations.
ISSN:1549-8328
1558-0806
DOI:10.1109/TCSI.2024.3419020