Video Experimental Relacionado
Updated: Sep 1, 2025

07:46
A Method for Growing Bio-memristors from Slime Mold
Published on: November 2, 2017
9.0K
Un chip de computación en memoria basado en memoria de acceso aleatorio resistivo
Weier Wan1,2, Rajkumar Kubendran3,4, Clemens Schaefer5
1Stanford University, Stanford, CA, USA. weierwan@stanford.edu.
Nature
|August 17, 2022
Resumen
Este estudio presenta NeuRRAM, un nuevo chip de cálculo en memoria (CIM) que utiliza memoria de acceso aleatorio resistivo (RRAM). NeuRRAM logra una eficiencia energética y una precisión superiores para las tareas de inteligencia artificial (IA) en dispositivos de borde.
Área de la Ciencia:
- Ciencias de los materiales
- Ingeniería informática
- Inteligencia artificial
Sus antecedentes:
- Los dispositivos de borde requieren hardware eficiente en energía para funcionalidades complejas de IA.
- La computación en memoria (CIM) utilizando memoria de acceso aleatorio resistivo (RRAM) ofrece una solución mediante la integración de memoria y computación.
- Los chips RRAM-CIM existentes se enfrentan a desafíos para equilibrar la eficiencia energética, la versatilidad del modelo y la precisión.
Objetivo del estudio:
- Desarrollar un chip CIM basado en RRAM, NeuRRAM, que supere las compensaciones entre eficiencia, versatilidad y precisión.
- Para demostrar mejoras simultáneas en múltiples jerarquías de diseño: algoritmos, arquitectura, circuitos y dispositivos.
- Habilitar funcionalidades avanzadas de IA directamente en dispositivos de borde con una eficiencia energética sin precedentes.
Principales métodos:
- Co-optimización entre algoritmos, arquitectura, circuitos y dispositivos para el diseño de RRAM-CIM.
- Desarrollo de un nuevo chip CIM basado en RRAM llamado NeuRRAM.
- Integración de dispositivos RRAM densos, analógicos y no volátiles para el cálculo en memoria.
Principales resultados:
- NeuRRAM logra una eficiencia energética dos veces mayor en comparación con los chips RRAM-CIM anteriores.
- El chip demuestra versatilidad al reconfigurar núcleos CIM para diversas arquitecturas de modelos de IA.
- La precisión de la inferencia es comparable a los modelos de software con cuantización de peso de cuatro bits en varias tareas de IA, incluida la clasificación de imágenes y el reconocimiento de voz.
Conclusiones:
- NeuRRAM representa un avance significativo en la tecnología CIM basada en RRAM.
- El enfoque de co-optimización aborda con éxito las compensaciones de eficiencia, versatilidad y precisión.
- Esta tecnología allana el camino para un procesamiento de IA altamente eficiente y preciso en dispositivos de borde.
Más Videos Relacionados
Videos de Conceptos Relacionados
MOS Capacitor
940
A Metal-Oxide-Semiconductor (MOS) capacitor is a fundamental structure used extensively in semiconductor device technology, particularly in the fabrication of integrated circuits and MOSFETs (metal-oxide-semiconductor field-effect transistors). The MOS capacitor consists of three layers: a metal gate, a dielectric oxide, and a semiconductor substrate.
The metal gate is typically made from highly conductive materials such as aluminum or polysilicon. Beneath the metal gate lies a thin layer of...
The metal gate is typically made from highly conductive materials such as aluminum or polysilicon. Beneath the metal gate lies a thin layer of...
940
Understanding Memory
603
Memory is the retention of information or experiences over time, facilitated through three main processes: encoding, storage, and retrieval. Encoding is the process of inputting information into the memory system. For instance, when listening to a lecture, watching a play, reading a book, or having a conversation, the brain is actively encoding information. This initial stage involves transforming sensory input into a form that can be processed and stored by the brain. Various factors, such as...
603
Non-ohmic Devices
1.2K
In most substances, the current flow is proportional to the voltage applied to it. A simple relationship between the values of current, voltage, and resistance is known as Ohm's law. Nonohmic devices do not exhibit a linear relationship between voltage and current. One such device is the semiconducting circuit element known as a diode. A diode is a circuit device that allows current flow in only one direction.
Consider a simple circuit consisting of a battery, a diode, and a resistor. A...
Consider a simple circuit consisting of a battery, a diode, and a resistor. A...
1.2K
System of Memory
6.4K
Memory is categorized into three major systems: sensory memory, short-term memory (STM), and long-term memory (LTM). These systems differ in their capacity and the duration for which they can hold information. Sensory memory captures raw sensory input from the environment, holding it for just a few seconds or less. For example, on hearing a brief, loud sound, like a car horn honking, the sound seems to linger in the mind for a moment even after it stops. This is an instance of sensory memory...
6.4K
Design Example: Resistive Touchscreen
415
A device engineer plays a crucial role in designing user interfaces for mobile devices. One such interface is the resistive touchscreen, which fundamentally consists of two metallic layers: a flexible upper layer and a rigid lower layer, separated by a narrow gap. The high resistance between these two layers is a key characteristic of this design.
When a user touches the screen, the two layers make contact at a specific point known as the touchpoint. This contact reduces the resistance between...
When a user touches the screen, the two layers make contact at a specific point known as the touchpoint. This contact reduces the resistance between...
415
Semiconductors
856
There is variation in the electrical conductivity of materials - metals, semiconductors, and insulators that are showcased with the help of the energy band diagrams.
Metals such as copper (Cu), zinc (Zn), or lead (Pb) have low resistivity and feature conduction bands that are either not fully occupied or overlap with the valence band, making a bandgap non-existent. This allows electrons in the highest energy levels of the valence band to easily transition to the conduction band upon gaining...
Metals such as copper (Cu), zinc (Zn), or lead (Pb) have low resistivity and feature conduction bands that are either not fully occupied or overlap with the valence band, making a bandgap non-existent. This allows electrons in the highest energy levels of the valence band to easily transition to the conduction band upon gaining...
856

