Non-Gaussian distribution of SRAM read current and design impact to low power memory using Voltage Acceleration Method

Author

Wang, Joseph ; Liu, Ping ; Gao, Yandong ; Deshmukh, Pankaj ; Yang, Sam ; Chen, Ying ; Sy, Wing ; Ge, Lixin ; Terzioglu, Esin ; Abu-Rahma, Mohamed ; Garg, Manish ; Yoon, Sei Seung ; Han, Michael ; Sani, Mehdi ; Yeap, Geoffrey

Author_Institution

Qualcomm Inc., San Diego, CA, USA

fYear

2011

fDate

14-16 June 2011

Firstpage

220

Lastpage

221

Abstract

SRAM read current tail distribution beyond 6s was studied using Voltage Acceleration Method (VAM). For the first time, non-Gaussian distribution of SRAM and ROM read current was confirmed with direct measurements on actual silicon. Data shows that conventional assumption of Gaussian distribution in read current is inaccurate especially at low Vdd and cold temperature conditions for low power memory in 28nm and beyond technology nodes. In 28nm, this inaccuracy would lead to 2× bit access delay penalty.

Keywords

SRAM chips; integrated circuit design; low-power electronics; read-only storage; silicon; ROM read current; SRAM read current tail distribution; VAM; design impact; low-power memory; nonGaussian distribution; silicon; size 28 nm; voltage acceleration method; Delay; Gaussian distribution; Random access memory; Read only memory; Silicon; Temperature measurement; Very large scale integration;

fLanguage

English

Publisher

ieee

Conference_Titel

VLSI Technology (VLSIT), 2011 Symposium on

Conference_Location

Honolulu, HI

ISSN

0743-1562

Print_ISBN

978-1-4244-9949-6

Type

conf

Filename

5984647