Related papers: A Novel Design for SRAM Bitcell with 3-Complementa…

DRAM-based main memory and its associated components increasingly account for a significant portion of application performance bottlenecks and power budget demands inside the computing ecosystem. To alleviate the problems of storage density…

Accelerating finite automata processing is critical for advancing real-time analytic in pattern matching, data mining, bioinformatics, intrusion detection, and machine learning. Recent in-memory automata accelerators leveraging SRAMs and…

To support emerging applications ranging from holographic communications to extended reality, next-generation mobile wireless communication systems require ultra-fast and energy-efficient baseband processors. Traditional complementary…

In this article, we present a configurable field-effect transistor (FET), where not only polarity (n- and p-type), but the conduction mechanism of a FET can also be configured dynamically. As a result, we can have both types of devices,…

Resistive Random Access Memory (RRAM) is an emerging device for processing-in-memory (PIM) architecture to accelerate convolutional neural network (CNN). However, due to the highly coupled crossbar structure in the RRAM array, it is…

We present a data-calibrated compact model of carbon nanotube (CNT) field-effect transistors (CNFETs) including contact resistance, direct source-to-drain and band-to-band tunneling currents. The model captures the effects of dimensional…

Designing lightweight convolutional neural network (CNN) models is an active research area in edge AI. Compute-in-memory (CIM) provides a new computing paradigm to alleviate time and energy consumption caused by data transfer in von Neumann…

The maturity of the chemical vapor deposition graphene-based device processing has increased from chip level demonstrations to wafer-scale fabrication in the past few years. Due to this wafer-scale, electrical characterization and analysis…

This paper summarizes the idea of Subarray-Level Parallelism (SALP) in DRAM, which was published in ISCA 2012, and examines the work's significance and future potential. Modern DRAMs have multiple banks to serve multiple memory requests in…

We present a simple and scalable technique for the fabrication of solution processed & local gated carbon nanotube field effect transistors (CNT-FETs). The approach is based on directed assembly of individual single wall carbon nanotube…

The current state of the art of Simultaneous Localisation and Mapping, or SLAM, on low power embedded systems is about sparse localisation and mapping with low resolution results in the name of efficiency. Meanwhile, research in this field…

We report strategies of achieving both high assembly yield of carbon nanotubes at selected position of the circuit via dielectrophoresis (DEP) and field effect transistor (FET) yield using semiconducting enriched single walled carbon…

The proliferation of Transformer models is often constrained by the significant computational and memory bandwidth demands of deployment. To address this, we present MXFormer, a novel, hybrid, weight-stationary Compute-in-Memory (CIM)…

Ion-Sensitive Field-Effect Transistors (ISFETs) form a wide-spread technology for sensing, thanks to their label-free detection and intrinsic CMOS compatibility. Their current sensitivity, {\Delta}ID/ID, for a given {\Delta}pH, however, is…

Tunnel Field Effect Transistors (TFET) have extremely low leakage current, exhibit excellent subthreshold swing and are less susceptible to short channel effects. However, TFETs do face certain special challenges, particularly with respect…

In this paper we present a comprehensive design and benchmarking study of Content Addressable Memory (CAM) at the 7nm technology node in the context of similarity search applications. We design CAM cells based on SRAM, spin-orbit torque,…

Neural networks are an increasingly attractive algorithm for natural language processing and pattern recognition. Deep networks with >50M parameters are made possible by modern GPU clusters operating at <50 pJ per op and more recently,…

Compute in-memory (CIM) is a promising technique that minimizes data transport, the primary performance bottleneck and energy cost of most data intensive applications. This has found wide-spread adoption in accelerating neural networks for…

We propose a co-design approach for compute-in-memory inference for deep neural networks (DNN). We use multiplication-free function approximators based on ell_1 norm along with a co-adapted processing array and compute flow. Using the…

Compute-in-memory (CiM) is a promising approach to improving the computing speed and energy efficiency in dataintensive applications. Beyond existing CiM techniques of bitwise logic-in-memory operations and dot product operations, this…