GSTDTAP

浏览/检索结果: 共4条,第1-4条 帮助

已选(0)清除 条数/页:   排序方式:
Classification with a disordered dopantatom network in silicon 期刊论文
NATURE, 2020, 577 (7790) : 341-+
作者:  Vagnozzi, Ronald J.;  Maillet, Marjorie;  Sargent, Michelle A.;  Khalil, Hadi;  Johansen, Anne Katrine Z.;  Schwanekamp, Jennifer A.;  York, Allen J.;  Huang, Vincent;  Nahrendorf, Matthias;  Sadayappan, Sakthivel;  Molkentin, Jeffery D.
收藏  |  浏览/下载:34/0  |  提交时间:2020/07/03

Classification is an important task at which both biological and artificial neural networks excel(1,2). In machine learning, nonlinear projection into a high-dimensional feature space can make data linearly separable(3,4), simplifying the classification of complex features. Such nonlinear projections are computationally expensive in conventional computers. A promising approach is to exploit physical materials systems that perform this nonlinear projection intrinsically, because of their high computational density(5), inherent parallelism and energy efficiency(6,7). However, existing approaches either rely on the systems'  time dynamics, which requires sequential data processing and therefore hinders parallel computation(5,6,8), or employ large materials systems that are difficult to scale up(7). Here we use a parallel, nanoscale approach inspired by filters in the brain(1) and artificial neural networks(2) to perform nonlinear classification and feature extraction. We exploit the nonlinearity of hopping conduction(9-11) through an electrically tunable network of boron dopant atoms in silicon, reconfiguring the network through artificial evolution to realize different computational functions. We first solve the canonical two-input binary classification problem, realizing all Boolean logic gates(12) up to room temperature, demonstrating nonlinear classification with the nanomaterial system. We then evolve our dopant network to realize feature filters(2) that can perform four-input binary classification on the Modified National Institute of Standards and Technology handwritten digit database. Implementation of our material-based filters substantially improves the classification accuracy over that of a linear classifier directly applied to the original data(13). Our results establish a paradigm of silicon-based electronics for smallfootprint and energy-efficient computation(14).


  
Fully hardware-implemented memristor convolutional neural network 期刊论文
NATURE, 2020, 577 (7792) : 641-+
作者:  Yoshioka-Kobayashi, Kumiko;  Matsumiya, Marina;  Niino, Yusuke;  Isomura, Akihiro;  Kori, Hiroshi;  Miyawaki, Atsushi;  Kageyama, Ryoichiro
收藏  |  浏览/下载:58/0  |  提交时间:2020/07/03

Memristor-enabled neuromorphic computing systems provide a fast and energy-efficient approach to training neural networks(1-4). However, convolutional neural networks (CNNs)-one of the most important models for image recognition(5)-have not yet been fully hardware-implemented using memristor crossbars, which are cross-point arrays with a memristor device at each intersection. Moreover, achieving software-comparable results is highly challenging owing to the poor yield, large variation and other non-ideal characteristics of devices(6-9). Here we report the fabrication of high-yield, high-performance and uniform memristor crossbar arrays for the implementation of CNNs, which integrate eight 2,048-cell memristor arrays to improve parallel-computing efficiency. In addition, we propose an effective hybrid-training method to adapt to device imperfections and improve the overall system performance. We built a five-layer memristor-based CNN to perform MNIST10 image recognition, and achieved a high accuracy of more than 96 per cent. In addition to parallel convolutions using different kernels with shared inputs, replication of multiple identical kernels in memristor arrays was demonstrated for processing different inputs in parallel. The memristor-based CNN neuromorphic system has an energy efficiency more than two orders of magnitude greater than that of state-of-the-art graphics-processing units, and is shown to be scalable to larger networks, such as residual neural networks. Our results are expected to enable a viable memristor-based non-von Neumann hardware solution for deep neural networks and edge computing.


  
Zucchini consensus motifs determine the mechanism of pre-piRNA production 期刊论文
NATURE, 2020, 578 (7794) : 311-+
作者:  Clark, Timothy D.;  Raby, Graham D.;  Roche, Dominique G.;  Binning, Sandra A.;  Speers-Roesch, Ben;  Jutfelt, Fredrik;  Sundin, Josefin
收藏  |  浏览/下载:34/0  |  提交时间:2020/07/03

PIWI-interacting RNAs (piRNAs) of between approximately 24 and 31 nucleotides in length guide PIWI proteins to silence transposons in animal gonads, thereby ensuring fertility(1). In the biogenesis of piRNAs, PIWI proteins are first loaded with 5 '  -monophosphorylated RNA fragments called pre-pre-piRNAs, which then undergo endonucleolytic cleavage to produce pre-piRNAs(1,2). Subsequently, the 3 '  -ends of pre-piRNAs are trimmed by the exonuclease Trimmer (PNLDC1 in mouse)(3-6) and 2 '  -O-methylated by the methyltransferase Hen1 (HENMT1 in mouse)(7-9), generating mature piRNAs. It is assumed that the endonuclease Zucchini (MitoPLD in mouse) is a major enzyme catalysing the cleavage of pre-pre-piRNAs into pre-piRNAs(10-13). However, direct evidence for this model is lacking, and how pre-piRNAs are generated remains unclear. Here, to analyse pre-piRNA production, we established a Trimmer-knockout silkworm cell line and derived a cell-free system that faithfully recapitulates Zucchini-mediated cleavage of PIWI-loaded pre-pre-piRNAs. We found that pre-piRNAs are generated by parallel Zucchini-dependent and -independent mechanisms. Cleavage by Zucchini occurs at previously unrecognized consensus motifs on pre-pre-piRNAs, requires the RNA helicase Armitage, and is accompanied by 2 '  -O-methylation of pre-piRNAs. By contrast, slicing of pre-pre-piRNAs with weak Zucchini motifs is achieved by downstream complementary piRNAs, producing pre-piRNAs without 2 '  -O-methylation. Regardless of the endonucleolytic mechanism, pre-piRNAs are matured by Trimmer and Hen1. Our findings highlight multiplexed processing of piRNA precursors that supports robust and flexible piRNA biogenesis.


A silkworm model recapitulates key steps of Zucchini-mediated cleavage of pre-pre-piRNA and provides insights into Zucchini-mediated and -independent pathways that generate pre-piRNAs, which converge to a common piRNA maturation step.


  
Final Report for "Analyzing and visualizing next generation climate data" 科技报告
来源:US Department of Energy (DOE). 出版年: 2012
作者:  Pletzer, Alexander
收藏  |  浏览/下载:14/0  |  提交时间:2019/04/05
Mosaic grids  parallel processing  block structured meshes  climate data  interpolation  regridding  UV-CDAT  ESMF