Design of Highly-Stable Energy-Minimum-Point SRAM Using Ultralow-Voltage Retention Cell
A highly-stable energy-minimum-point (EMP) SRAM cell ($10 \text{T}_{\text{IRP}}^{\text{E} / \text{U} , \text{C} / \text{S}}$) is proposed, which is configured with an ultralow-voltage-retention (ULVR) 8T<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">U,C/C</sup> cell and an isolated-read-port (IRP), i.e., $10 \text{T}_{\text{IRP}}^{\text{E} / \text{U} , \text{C} / \text{S}} = 8 \text{T}^{\text{U} , \text{C} / \text{C}} + \text{IRP}$ (in which the suffix C/C represents the usage of complementary signals for the write and read operations, and C/S represents complementary-signal write and single-ended read operations). The $10 \text{T}_{\text{IRP}}^{\text{E} / \text{U} , \text{C} / \text{S}}$ cell can achieve not only stable ULVR but also energy-efficient EMP operations (the suffix E/U shows these two abilities). The $10 \text{T}_{\text{IRP}}^{\text{E} / \text{U} , \text{C} / \text{S}}$ cell is optimally designed so that the powerful feedback effect of the 8T<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">U,C/C</sup> cell appears even at EMP, resulting in ~2 times higher noise margin than our previously-developed EMP/ULVR-cell (10T<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">E/U,C/C</sup>). An 8KB EMP/ULVR-SRAM macro using a $10 \text{T}_{\text{IRP}}^{\text{E} / \text{U} , \text{C} / \text{S}}$ cell array is also designed. The EMP and ULVR operations can reduce the active and standby power by ~98%, respectively, from the normal voltage operation mode.
Authors
- Yusaku Shiotsu (ORCID: https://orcid.org/0000-0002-3586-9772)
- Satoshi Sugahara
- Katsutoshi Ito
Institutions
- Tokyo Institute of Technology (JP)
Publication Details
- Published
- 2024-11-21
- DOI
- https://doi.org/10.1109/imfedk64776.2024.10814372
- Primary Topic
- Low-power high-performance VLSI design
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- Synopsys
- University of Tokyo