English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Computing in-memory (CiM) has emerged as an attractive technique to mitigatethe von-Neumann bottleneck. Current digital CiM approaches for in-memoryoperands are based on multi-wordline assertion for computing bit-wise Booleanfunctions and arithmetic functions such as addition. However, most of thesetechniques, due to the many-to-one mapping of input vectors to bitlinevoltages, are limited to CiM of commutative functions, leaving out an importantclass of computations such as subtraction. In this paper, we propose a CiMapproach, which solves the mapping problem through an asymmetric wordlinebiasing scheme, enabling (a) simultaneous single-cycle memory read and CiM ofprimitive Boolean functions (b) computation of any Boolean function and (c) CiMof non-commutative functions such as subtraction and comparison. While theproposed technique is technology-agnostic, we show its utility forferroelectric transistor (FeFET)-based non-volatile memory. Compared to thestandard near-memory methods (which require two full memory accesses peroperation), we show that our method can achieve a full scale two-operanddigital CiM using just one memory access, leading to a 23.2% - 72.6% decreasein energy-delay product (EDP).

ADRA: Extending Digital Computing-in-Memory with Asymmetric Dual-Row-Activation