International Business Machines Corporation

Globalfoundries Inc.

King Abdulaziz City for Science and Technology

Elpis Technologies Inc.

Bahman Hekmatshoartabari

Davood Shahrjerdi

Alexander Reznicek

Ghavam G Shahidi

Devendra K Sadana

Stephen W Bedell

Karthik Balakrishnan

Tak H Ning

Ali Afzali-Ardakani

Tze-Chiang Chen

Jeng-Bang Yau

Ali Khakifirooz

Pouya Hashemi

Takashi Ando

Kevin K Chan

Kailash Gopalakrishnan

Anirban Basu

Ruilong Xie

Ning Li

Joel Pereira De Souza

Nanbo Gong

Wilfried Haensch

Fabio Carta

Keith E Fogel

Jeehwan Kim

Matthew Joseph BrightSky

Daniel Michael Kuchta

Chung H Lam

Jingyun Zhang

Jeremy R Fox

Martin G Keen

Heng Wu

Katherine Lynn Saenger

Stephen McConnell Gates

Siegfried Lutz Maurer

Clint Oteri

Daniel C Edelstein

Ravi Nair

John Albrecht Ott

Maryam Ashoori

Young Hoon Kwark

Elham Khabiri

Wilfried Ernst-August Haensch

Suyog Gupta

Warren S Rieutort-Louis

Homa Alemzadeh

Oleg Gluschenkov

Praneet Adusumilli

Oscar Van Der Straten

Paul Alfred Lauro

Cheng-Wei Cheng

Oki Gunawan

Jin Cai

Dae-Gyu Park

Robert L Bruce

Yanning Sun

Christopher Peter D'Emic

Arjang Hassibi

Yasir Sulehria

Abdulrahman Albadri

Marinus P J Hopstaken

Choonghyun Lee

Michael Rizzolo

Kenneth Parker Rodbell

Dirk Pfeiffer

Talia S Gershon

Harold John Hovel

Gregory M Fritz

Tsung-Sheng Kang

Wanki Kim

Liying Jiang

ChoongHyun Lee

Eric Raymond Evarts

Asit Kumar Ray

Virat Vasav Mehta

Marinus Pj Hopstaken

Davood Shahjerdi

Training machine learning systems using a training data set, gradient descent, and a loss function. The machine learning system includes memory and reads and writes to memory according to read and write profiles. The loss function is associated with machine learning system memory read and write profile gradients. The loss function includes a loss function penalty term, the loss function penalty term being associated with the read and write profile gradient differences. Trained machine learning systems are then provided.

Training machine learning systems

A sensor device includes a vertically stacked cascode bipolar junction transistor pair, and a first trench having a first sidewall, wherein a portion of the first sidewall is provided by the first sensing surface, wherein a bipolar junction transistor and a dual-base bipolar junction transistor of the cascode bipolar junction transistor pair are stacked vertically along the first trench.

Vertically stacked cascode bipolar junction transistor (BJT) pair sensor

A semiconductor tunnel FET (field effect transistor) including a plurality of nanosheet channels disposed between a first source/drain region and a second source/drain region. The first source/drain region includes a p-type material; and the second source/drain region includes an n-type material.

Tunnel field effect transistor devices

A hybrid stacked semiconductor device includes a nanosheet stack on a substrate and an all-around gate. The nanosheet stack includes a first stack portion and a second stack portion. The first stack portion includes first channels. The second stack portion is stacked on the first stack portion, and includes second channels. The all-around gate includes a first gate portion that wraps around the first channels and a second gate portion that wraps around the second channels. A first gate extension contacts the first gate portion and the second gate extension contacts the second gate portion. At least one gate contact contacts the first gate extension to establish conductivity with the first gate portion and contacts the second gate extension to establish conductivity with the second gate portion.

Hybrid stacked field effect transistors

Embodiments of the present invention include a phase change memory (PCM) array. The PCM array may include a plurality of PCM cells. Each PCM cell in the plurality of PCM cells may include a top electrode, a resistive element, and a bottom electrode. The PCM array may also include a global heater surrounding the plurality of PCM cells having a thermally conductive material contacting each of the plurality of PCM cells. The global heater may be configured to receive an electric signal to heat the plurality of PCM cells simultaneously.

Global heater for phase change memory

A bottom electrode, a phase change material layer, the phase change material layer includes a similar lattice constant as a lattice constant of the bottom electrode, and a top electrode vertically aligned. A phase change material layer, a top electrode adjacent to a first vertical side surface of the phase change material layer, and a bottom electrode adjacent to a second vertical side surface of the phase change material layer. Forming a phase change material layer, forming a top electrode adjacent to a first vertical side surface and overlapping a first portion of an upper horizontal surface of the phase change material layer, forming a bottom electrode, adjacent to a second vertical side surface and overlapping a second portion of the upper horizontal surface of the phase change material layer, and forming a dielectric material horizontally isolating the bottom electrode and the top electrode.

Crossbar memory array in back end of line

A crystallization seed layer in a substrate, a phase change material layer, the phase change material layer includes a similar lattice constant as a lattice constant of the crystallization seed layer, a top electrode adjacent to a first vertical side surface and a bottom electrode adjacent to a second vertical side surface of the phase change material layer. A plurality of memory structures configured in a crossbar array, each including a crystallization seed layer, a phase change material layer above, a top electrode adjacent to a first vertical side surface and a bottom electrode adjacent to a second vertical side surface of the phase change material layer. A method including forming a crystallization seed layer, forming a phase change material layer, forming a top electrode and a bottom electrode on the substrate, each adjacent to a vertical side surface of the phase change material layer.

Crossbar memory array in back end of line with crystallization front

Embodiments of the invention are directed to a computer-implemented method that includes accessing, using a processor system, a three-dimensional (3D) model of a device-under-design (DUD). The processor system is used to receive a first design operation associated with the 3D model of the DUD. A collaboration dependency model is used to make a prediction of a dependency relationship between the first design operation and a second design operation.

Virtual reality design navigation using a temporal collaboration dependency map

Methods, systems, and a computer program product are disclosed. The first method includes obtaining virtual session data in real time, identifying a positional utterance in the virtual session data, and generating a positional insight for the positional utterance. The first method also includes generating a best-practices recommendation based on the positional insight. The second method includes obtaining virtual session data, identifying positional utterances in the virtual session data, and generating positional insights for each of the positional utterances. The second method also includes selecting each of the positional insights having confidence scores above a threshold score and generating best-practices recommendations based on the selected positional insights.

Contextual positioning in virtual space

Methods, systems, and a computer program product are disclosed. The first method includes obtaining virtual session data in real time, identifying a positional utterance in the virtual session data, and generating a positional insight for the positional utterance. The first method also includes rendering a user avatar in a position recommended based on the positional insight. The second method includes obtaining virtual session data in real time, identifying a positional utterance in the virtual session data, and generating positional insights for the positional utterance. The second method also includes generating at least one position recommendation based on the positional insights.

A Darlington pair sensor is disclosed. The Darlington pair sensor has an amplifying/horizontal bipolar junction transistor (BJT) and a sensing/vertical BJT and can be used as a biosensor. The amplifying bipolar junction transistor (BJT) is horizontally disposed on a substrate. The amplifying BJT has a horizontal emitter, a horizontal base, a horizontal collector, and a common extrinsic base/collector. The common extrinsic base/collector is an extrinsic base for the amplifying BJT. The sensing BJT has a vertical orientation with respect to the amplifying BJT. The sensing BJT has a vertical emitter, a vertical base, an extrinsic vertical base, and the common extrinsic base/collector (in common with the amplifying BJT). The common extrinsic base/collector acts as the sensing BJT collector. The extrinsic vertical base is separated into a left extrinsic vertical base and a right extrinsic vertical base giving the sensing BJT has two separated (dual) bases, a sensing base and a control base. The Darlington pair sensor has high in-situ signal amplification with low noise and uses substrate space effectively.

Darlington pair bipolar junction transistor sensor

A spin-orbit torque (SOT)-MRAM comprising a first magnetic tunneling junction (MTJ) having a first distance and having a first critical voltage. A second MTJ having a second distance and having a second critical voltage, wherein the first distance and the second distance are different, wherein the first critical voltage and the second critical voltages are different. A metal rail in direct contact with the first MTJ and the second MTJ, wherein the metal rail injects a spin current in to both the first MTJ and the second MTJ.

Multi-state SOT-MRAM structure

In an approach to improve the generation of a virtual object in a three-dimensional virtual environment, embodiments of the present invention identify a virtual object to be generated in a three-dimensional virtual environment based on a natural language utterance. Additionally, embodiments generate the virtual object based on a CLIP-guided Generative Latent Space (CLIP-GLS) analysis, and monitor usage of the generated virtual object in the three-dimensional virtual space. Moreover, embodiments infer human perception data from the monitoring, and generate a utility score for the virtual object based on the human perception data.

Optimizing computer-based generation of three-dimensional virtual objects

A circuit structure includes a first ferroelectric field effect transistor (FeFET) having a first gate electrode, a first source electrode, and a first drain electrode and a second FeFET having a second gate electrode, a second source electrode, and a second drain electrode. The first gate electrode is connected to a wordline, and the first source electrode and the second source electrode are connected to a bitline. The first drain electrode is connected to the second gate electrode and the second drain electrode is connected to a bias line. A weight synapse structure is constructed by combining two circuit structures. A plurality of weight synapse structures are incorporated into a crossbar array.

FeFET unit cells for neuromorphic computing

Embodiments disclosed herein include a semiconductor structure. The semiconductor structure may include a spin transfer torque (STT) magnetoresistive random access memory (MRAM) stack. The semiconductor structure may also include a spin orbit torque (SOT) MRAM stack vertically in series with the STT-MRAM. The SOT-MRAM stack may include a heavy metal spin hall effect rail configured to flip an SOT free-layer magnetic orientation in response to a horizontal signal through the heavy metal rail.

Spin-orbit torque and spin-transfer torque magnetoresistive random-access memory stack

Growing community of inventors

White Plains, NY, United States of America

Bahman Hekmatshoartabari