Intel Corporation

Tahoe Research, Ltd.

Omek Interactive, Inc.

Amit Bleiweiss

Yaniv Fais

Lev Faivishevsky

Jacob Subag

Tomer Bar-On

Jeremie Dreyfuss

Tomer Schwartz

Itamar Ben-Ari

Michael Behar

Uzi Sarel

Amitai Armon

Gal Leibovich

Ehud Cohen

Yahav Shadmiy

Guy Jacob

Eriko Nurvitadhi

Gershom Kutliroff

Yaron Yanai

Deborah T Marr

Dhawal Srivastava

Galina Ryvchin

Ofir Levy

Eyal Yaacoby

Chen Paz

Elmoustapha Ould-Ahmed-Vall

Mattias Marder

Jeffrey J Cook

Srinivas Sridharan

Asit K Mishra

Dheevatsa Mudigere

Raanan Yonatan Yehezkel Rohekar

Anavai Ramesh

Mohammad Ashraf Bhuiyan

Estelle Aflalo

Sabareesh Ganapathy

Md Faijul Amin

Eliran Zimmerman

Saritha Dwarakapuram

Eugene Wang

Niharika Maheshwari

Shahar Fleishman

Eran Ben-Avi

Neta Zmora

Jonathan Epstein

Yotam Livny

Itamar Glazer

Maoz Madmoni

Travis T Schluessler

Wei W Wang

Abhishek Venkatesh

Devan Burke

Wei D Wang

Wei Wang

John G Gierach

Oguz Hasan Elibol

Eran Eilat

Anil Thomas

Gokce Keskin

Amit Gur

Haim Barad

Eran Dagan

Anat Heilper

Yinon Oshrat

Dagan Eshar

Amir Rosenberger

Pruthvi Gowda Thorehosur Appajigowda

Nishant Patel

Huma Abidi

Jaikrishnan Menon

Barak Battach

Amital Armon

An apparatus to facilitate acceleration of machine learning operations is disclosed. The apparatus comprises at least one processor to perform operations to implement a neural network and accelerator logic to perform communicatively coupled to the processor to perform compute operations for the neural network.

Machine learning accelerator mechanism

An apparatus to facilitate processing of a sparse matrix for arbitrary graph data is disclosed. The apparatus includes a graphics processing unit having a data management unit (DMU) that includes a scheduler for scheduling matrix operations, an active logic for tracking active input operands, and a skip logic for tracking unimportant input operands to be skipped by the scheduler. Processing circuitry is coupled to the DMU. The processing circuitry comprises a plurality of processing elements including logic to read operands and a multiplication unit to multiply two or more operands for the arbitrary graph data and customizable circuitry to provide custom functions.

Machine learning sparse computation mechanism for arbitrary neural networks, arithmetic compute microarchitecture, and sparsity for training mechanism

Methods, apparatus, systems and articles of manufacture are disclosed to map workloads. An example apparatus includes a constraint definer to define performance characteristic targets of the neural network, an action determiner to apply a first resource configuration to candidate resources corresponding to the neural network, a reward determiner to calculate a results metric based on (a) resource performance metrics and (b) the performance characteristic targets, and a layer map generator to generate a resource mapping file, the mapping file including respective resource assignments for respective corresponding layers of the neural network, the resource assignments selected based on the results metric.

Methods, systems, articles of manufacture and apparatus to map workloads

In an example, an apparatus comprises a plurality of execution units comprising and logic, at least partially including hardware logic, to receive a plurality of data inputs for training a neural network, wherein the data inputs comprise training data and weights inputs; represent the data inputs in a first form; and represent the weight inputs in a second form. Other embodiments are also disclosed and claimed.

Real time context dependent deep learning

Methods, apparatus, systems, and articles of manufacture are disclosed that optimize layers of a machine learning model for a target hardware platform. An example apparatus includes a communication processor to obtain information specific to the target hardware platform (THP) on which to execute the machine learning model; a layer generation controller to generate layers of the machine learning model based on the information specific to the THP; and a deployment controller to, in response to the machine learning model satisfying a threshold error metric, deploy the machine learning model to the THP.

Methods, systems, articles of manufacture, and apparatus to optimize layers of a machine learning model for a target hardware platform

An apparatus to facilitate compute optimization is disclosed. The apparatus includes a at least one processor to perform operations to implement a neural network and compute logic to accelerate neural network computations.

Compute optimization mechanism for deep neural networks

Techniques are provided for generation of synthetic 3-dimensional object image variations for training of recognition systems. An example system may include an image synthesizing circuit configured to synthesize a 3D image of the object (including color and depth image pairs) based on a 3D model. The system may also include a background scene generator circuit configured to generate a background for each of the rendered image variations. The system may further include an image pose adjustment circuit configured to adjust the orientation and translation of the object for each of the variations. The system may further include an illumination and visual effect adjustment circuit configured to adjust illumination of the object and the background for each of the variations, and to further adjust visual effects of the object and the background for each of the variations based on application of simulated camera parameters.

Generation of synthetic 3-dimensional object images for recognition systems

An activity recording system is provided. The activity recording system includes a three-dimensional camera, a sensor arrangement that is fitted to a subject being recorded, and an activity recording device. The activity recording device receives image information from the three-dimensional camera and sensor arrangement information from the sensor arrangement. Both the image information and the sensor arrangement information include location measurements. The sensor arrangement information is generated by location sensors that are positioned at target features of the subject to be tracked. The sensor arrangement information is a key to the image information that specifies where, in any given image, the target features of the subject lie. Activity data having these characteristics may be applied to solve a variety of system development problems. Such activity data can be used to training machine learning components or test computer vision components for a fraction of the cost of using conventional techniques.

Video feature tagging

In an example, an apparatus comprises a plurality of execution units comprising at least a first type of execution unit and a second type of execution unit and logic, at least partially including hardware logic, to expose embedded cast operations in at least one of a load instruction or a store instruction; determine a target precision level for the cast operations; and load the cast operations at the target precision level. Other embodiments are also disclosed and claimed.

Variable precision and mix type representation of multiple layers in a network

A mechanism is described for facilitating memory handling and data management in machine learning at autonomous machines. A method of embodiments, as described herein, includes detecting multiple tables associated with multiple neural networks at multiple autonomous machines, where each of the multiple tables include an index. The method may further include combining the multiple tables and multiple indexes associated with the multiple tables into a single table and a single index, respectively, where the single table is communicated to the multiple autonomous machines to allow simultaneous processing of one or more portions of the single table using one or more memory devices and one or more processors of one or more of the multiple autonomous machines.

Smart memory handling and data management for machine learning networks

In an example, an apparatus comprises logic, at least partially including hardware logic, to save one or more outputs of a deep learning neural network in a storage system of an autonomous vehicle and upload the one or more outputs to a remote server. Other embodiments are also disclosed and claimed.

Storage system of DNN outputs for black box

Growing community of inventors

Yad Binyamin, Israel

Amit Bleiweiss