Compressed Sensing With Approximate Message Passing Using In-Memory Computing

Gallo, Manuel Le; Sebastian, Abu; Cherubini, Giovanni; Giefers, Heiner; Eleftheriou, Evangelos

doi:10.1109/ted.2018.2865352

Cited by 82 publications

(62 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…3b. For all levels, we achieve a standard deviation less than 1.2 μS, which is more than two times lower than that reported in previous works on nanoscale PCM arrays for a similar conductance range 40,41 .…”

Section: Resnet Block 1 10 Layerscontrasting

confidence: 53%

Accurate deep neural network inference using computational phase-change memory

Joshi¹,

Gallo²,

Haefeli³

et al. 2020

Nat Commun

Self Cite

339

323

View full text Add to dashboard Cite

In-memory computing using resistive memory devices is a promising non-von Neumann approach for making energy-efficient deep learning inference hardware. However, due to device variability and noise, the network needs to be trained in a specific way so that transferring the digitally trained weights to the analog resistive memory devices will not result in significant loss of accuracy. Here, we introduce a methodology to train ResNet-type convolutional neural networks that results in no appreciable accuracy loss when transferring weights to phase-change memory (PCM) devices. We also propose a compensation technique that exploits the batch normalization parameters to improve the accuracy retention over time. We achieve a classification accuracy of 93.7% on CIFAR-10 and a top-1 accuracy of 71.6% on ImageNet benchmarks after mapping the trained weights to PCM. Our hardware results on CIFAR-10 with ResNet-32 demonstrate an accuracy above 93.5% retained over a one-day period, where each of the 361,722 synaptic weights is programmed on just two PCM devices organized in a differential configuration.

show abstract

Section: Resnet Block 1 10 Layerscontrasting

confidence: 53%

Accurate deep neural network inference using computational phase-change memory

Joshi¹,

Gallo²,

Haefeli³

et al. 2020

Nat Commun

Self Cite

339

323

View full text Add to dashboard Cite

show abstract

“…The simulator is a cycle-accurate model that imitates the functionality of the memristor computations and surrounding digital blocks [14]. The memristor crossbar is an 8-bit 256x256 PCM crossbar based on IBM's 4-bit PCM [4]. To mimic an 8-bit cell with a 4-bit cell, two adjacent columns are used, one for 4 MSBs and the other for 4 LSBs.…”

Section: Demonstration and Evaluationmentioning

confidence: 99%

TDO-CIM: Transparent Detection and Offloading for Computation In-memory

Vadivel

Chelini

BanaGozar

et al. 2020

2020 Design, Automation &Amp; Test in Europe Conference &Amp; Exhibition (DATE)

View full text Add to dashboard Cite

Computation in-memory is a promising non-von Neumann approach aiming at completely diminishing the data transfer to and from the memory subsystem. Although a lot of architectures have been proposed, compiler support for such architectures is still lagging behind. In this paper, we close this gap by proposing an end-to-end compilation flow for in-memory computing based on the LLVM compiler infrastructure. Starting from sequential code, our approach automatically detects, optimizes, and offloads kernels suitable for in-memory acceleration. We demonstrate our compiler tool-flow on the PolyBench/C benchmark suite and evaluate the benefits of our proposed inmemory architecture simulated in Gem5 by comparing it with a state-of-the-art von Neumann architecture.

show abstract

“…One possible mapping method is an iterative program-andverify procedure [47]. After the matrix A is programmed in the crossbar array, the measurements y can obtained through a vector-matrix multiplication as illustrated in Fig.…”

Section: Compressed Sensing and Recoverymentioning

confidence: 99%

The Power of Computation-in-Memory Based on Memristive Devices

Lebdeh

Nguyen

et al. 2020

2020 25th Asia and South Pacific Design Automation Conference (ASP-DAC)

View full text Add to dashboard Cite

Conventional computing architectures and the CMOS technology that they are based on are facing major challenges such as the memory bottleneck making the memory access for data transfer a major killer of energy and performance. Computation-in-memory (CIM) paradigm is seen as a potential alternative that could alleviate such problems by adding computational resources to the memory, and significantly reducing the communication. Memristive devices are promising enablers of a such CIM paradigm, as they are able to support both storage and computing. This paper shows the power of memristive device based CIM paradigm in enabling new efficient applicationspecific architectures as well as efficient implementations of some known domain-specific architectures. In addition, the paper discusses the potential applications that could benefit from such paradigm and highlights the major challenges.

show abstract

Compressed Sensing With Approximate Message Passing Using In-Memory Computing

Cited by 82 publications

References 21 publications

Accurate deep neural network inference using computational phase-change memory

Accurate deep neural network inference using computational phase-change memory

TDO-CIM: Transparent Detection and Offloading for Computation In-memory

The Power of Computation-in-Memory Based on Memristive Devices

Contact Info

Product

Resources

About