A modular and scalable architecture for the realization of high-speed programmable rank-order filters using threshold logic

Chattez avec Graph Search

Posez n’importe quelle question sur les cours, conférences, exercices, recherches, actualités, etc. de l’EPFL ou essayez les exemples de questions ci-dessous.

AVERTISSEMENT : Le chatbot Graph n'est pas programmé pour fournir des réponses explicites ou catégoriques à vos questions. Il transforme plutôt vos questions en demandes API qui sont distribuées aux différents services informatiques officiellement administrés par l'EPFL. Son but est uniquement de collecter et de recommander des références pertinentes à des contenus que vous pouvez explorer pour vous aider à répondre à vos questions.

Grokking is proposed and widely studied as an intricate phenomenon in which generalization is achieved after a long-lasting period of overfitting. In this work, we propose NeuralGrok, a novel gradient-based approach that learns an optimal gradient transformation to accelerate the generalization of transformers in arithmetic tasks. Specifically, NeuralGrok trains an auxiliary module (e.g., an MLP block) in conjunction with the base model. This module dynamically modulates the influence of individual gradient

Solid-state quantum computers require classical electronics to control and readout individual qubits and to enable fast classical data processing [1-3]. Integrating both subsystems at deep cryogenic temperatures [4], where solid-state quantum processors operate best, may solve some major scaling challenges, such as system size and input/output (I/O) data management [5]. Spin qubits in silicon quantum dots (QDs) could be monolithically integrated with complementary metal-oxide-semiconductor (CMOS) electronic

A novel platform suitable for fundamental investigations of light propagation through micro-and nanostructures will be discussed. This platform is based on a dielectric multilayer that sustains Bloch surface waves (BSWs). BSWs are electromagnetic surface waves excited at the interface between a truncated periodic dielectric multilayer and a surrounding media. The modification of the top surface to customize a complete 2D micro-system can be produced using e-beam writing, optical lithography or other pattern

A systematic comparison of front- and rear-emitter silicon heterojunction solar cells produced on nand p-type wafers was performed, in order to investigate their potential and limitations for high efficiencies. Cells on p-type wafers suffer from reduced minority carrier lifetime in the low-carrier-injection range, mainly due to the asymmetry in interface defect capture cross sections. This leads to slightly lower fill factors than for n-type cells. However, these losses can be minimized by using high-qualit

We derive scaling laws for the Rayleigh-Sommerfeld formulation we recently developed to describe light scattering from nanotextured interfaces. These scaling laws provide precious intuition on how to link scattering from different interfaces. In particular, we answer the question how to obtain a Lambertian scatterer into silicon, starting from a Lambertian scatterer into air relevant to the development of light trapping schemes in thin-film silicon solar cells. We also define a Lambertionality factor which

Look-Up Tables (LUTs) are universally used in FPGAs as the elementary logic blocks. They can implement any logic function and thus covering a circuit is a relatively straightforward problem. Naturally, flexibility comes at a price, and increasing the number of LUT inputs to cover larger parts of a circuit has an exponential cost in the LUT complexity. Hence, rarely LUTs with more than 4-6 inputs have been used. In this paper we argue that other elementary logic blocks can provide a better compromise between

An integrated chip for DNA hybridization detection was realized in a standard CMOS process: it hosts 80 biosensors subdivided in 2 channels, as well as D/A and A/D converters for electrical stimulation and readout. A microfluidic system, bonded on the surface of the chip, provides access to the sensing areas for test solutions and reagents. A simple, low-cost technique for the realization of a biocompatible interface on the aluminum sensing area was developed; it allows the immobilization of single-stranded

Present-day advanced technologies heavily rely on the exciting magnetic and spectroscopic properties of lanthanide ions. In particular, their ability to generate well-characterized and intense near-infrared (NIR) luminescence is exploited in any modern fiber-optic telecommunication network. In this feature article, we first summarize the whereabouts underlying the design of highly luminescent NIR molecular edifices and materials. We then focus on describing the main trends in three applications related to t

The electric field E within the i-layer of hydrogenated amorphous silicon (a-Si:H) solar cells strongly affects the cell performances, and, specifically, the fill factor FF. It governs the drift length Ldrift = μτE which is the crucial parameter limiting charge collection. Ideally, a constant electric field is assumed across the i-layer, whereas in real devices, it is deformed by charged band tail states and dangling bonds. If the i-layer is too thick or has a high density of charged defects, E is deformed

The paper presents an influence of leakage effect observed in capacitive analog memories on learning process in hardware implemented Kohonen neural networks with MOS transistors used as switches connected with information holding capacitors. The learning results, i.e. variations (adaptations) of weight values, strongly depend on the transistor leakage currents. This is a cause for some quantization error associated with the weight adaptations during the network training. The unwanted leakage influence can b

Harmonic probabilistic models are common in signal analysis. Framed as a linear-Gaussian state-space model, smoothed inference scales as

O(TH^2)

where

H

is twice the number of frequencies in the model and

T

is the length of the time-series. Due to their central role in acoustic modelling, fast effective inference in this model is of some considerable interest. We present a form of `rotation-corrected' low-rank approximation for the backward pass of the Rauch-Tung-Striebel smoother. This provides an ef

We study the interfacial electronic properties of a model Si-SiO2-Si structure which is intended to simulate the substrate-oxide-polysilicon stack in metal-oxide-semiconductor devices. The structural properties of this model are shown to match closely those of the real system, as determined by a number of atomic-scale experimental probes. In particular, the model introduced here takes into account the disordered reconstruction at the Si-SiO2 interface, and the amorphous nature of the 2 nm thick oxide. For t

In this paper an approach for localization using geometric features from a 360 degree laser range finder and a monocular vision system is presented. Its practicability under conditions of continuous localization during motion in real-time (referred to as on-the-fly localization) is investigated in large-scale experiments. The features are infinite horizontal lines for the laser and vertical lines for the camera. They are extracted using physically well-grounded models for all sensors and passed to a Kalman

We present a fully integrated Mach-Zehnder interferometer in silicon-on- insulator technology. Modulation of the index of refraction is achieved through the plasma dispersion effect resulting in a bandwidth in the 10 MHz range. A particular and innovative design makes this device completely compatible with CMOS technology allowing electronic functions to be integrated on the same substrate. Measurement results, limitations due to thermooptic effect and absorption related to charge injection together with fu

We present an algorithm for nonlinear multidimensional registration. The correspondence function is represented in a spline space, as well as the volumes being registered. We also use the cubic splines to interpolate the volumetric data to be registered. We use a multiresolution strategy, which gives us robustness and additional speedup. We analyze the computational complexity of the algorithm and its performance using different multidimensional optimization methods. Finally, we present an application of th

This paper suggests several approaches for determining the transient stability of a power network by using an analog VLSI chip for simulating the system behavior. The main advantages of using this method are the much shorter computation time and lower complexity compared to the existing methods that are based on numerical calculations or discrete analog emulators

A novel architecture descending from the work of von Neumann, has been developed. This architecture borrows its main principles from living systems. Like living beings, the organisms considered here are able to autonomously develop, maintain their functionality and reproduce. These genomic architectures are developed on reprogrammable hardware. They are not restricted to a given class of functions but accept any combinational and sequential function to be downloaded. These architectures are fault tolerant b

Stable negative ions containing up to sixteen silicon atoms have been measured by mass spectrometry in RF power-modulated silane plasmas for amorphous silicon deposition. These hydrogenated silicon cluster ions reach much higher masses than the positive ions, which have no more than six silicon atoms. This supports the view that negative ions are the precursors to particulate formation in silane plasmas. The time-dependent fluxes of positive and negative ions from the plasma are shown with a 5 mus time reso

Laser induced photolytic and pyrolytic deposition spots on silicon are studied for different laser powers, partial pressure and deposition time. In the photolytic case at low powers, the reaction takes place in the adsorbed phase on the surface, whereas at higher powers it is a predominantly gas phase process. Schottky diodes have been tested to demonstrate the good quality of platinum deposits.

A modular and scalable architecture for the realization of high-speed programmable rank-order filters using threshold logic

Chattez avec Graph Search

NeuralGrok: Accelerate Grokking by Neural Gradient Transformation

Integrated multiplexed microwave readout of silicon quantum dots in a cryogenic CMOS chip

Bloch surface waves, a 2D platform for planar optical integration

Silicon heterojunction solar cells on n- and p-type wafers with efficiencies above 20%

Light trapping in solar cells: When does a Lambertian scatterer scatter Lambertianly?

Rethinking FPGAs: Elude the Flexibility Excess of LUTs with And-Inverter Cones

Label-free, direct DNA detection by means of a standard CMOS electronic chip

Lanthanide NIR luminescence for telecommunications, bioanalyses and solar energy conversion

Internal electric field and fill factor of amorphous silicon (a-Si:H) solar cells

Influence of Information Leakage in Analog Memory on Learning Kohonen Network on Silicon

Efficient Kalman Smoothing for Harmonic State-Space Models

Electronic structure at realistic Si(100)-SiO2 interfaces

Multisensor On-the-Fly Localization: Precision and Reliability for Applications

CMOS compatible fully integrated Mach-Zehnder interferometer in SOI technology

Multidimensional Elastic Registration of Images Using Splines

On-chip transient stability simulator, 104 times faster than real time

Embryonics: The Birth of Synthetic Life

Negative Hydrogenated Silicon Ion Clusters as Particle Precursors in RF Silane Plasma Deposition Experiments

Platinum spots on silicon deposited by photon induced chemical vapor deposition