Intel Corporation

Jeremy Bottleson

Altug Koker

Eriko Nurvitadhi

Nadathur Rajagopalan Satish

Farshad Akhbari

Xiaoming Chen

Liwei Ma

Abhishek R Appu

Joydeep Ray

Elmoustapha Ould-Ahmed-Vall

Justin E Gottschlich

Anbang Yao

Kevin Nealis

Balaji Vembu

Narayan Srinivasa

Kamal Sinha

Feng Chen

Dukhwan Kim

Michael S Strickland

Sara S Baghsorkhi

Tatiana Shpeisman

Ping Tak Tang

Barath Lakshmanan

Jingyi Jin

Nicolas C Galoppo Von Borries

Rajkishore Barik

Tsung-Han Lin

Chandrasekaran Sakthivel

Vasanth Ranganathan

Sanjeev S Jahagirdar

Ben J Ashbaugh

Mike B Macpherson

Dhawal Srivastava

Kylan Race

Ernesto Zamora Ramos

Prasoonkumar Surti

Travis T Schluessler

Ankur N Shah

John Charles Weast

Jonathan Kennedy

Bradley H Smith

Michael Strikland

Alfredo Gimenez

An improved AR/VR operation includes receiving, by a server computing device, encrypted AR/VR user data and cleartext metadata associated with the encrypted AR/VR user data from a client computing device; getting server data based at least in part on cleartext metadata; encoding the server data; performing an AR/VR process on the encrypted AR/VR user data and the encoded server data to generate encrypted AR/VR results; and sending the encrypted AR/VR results to the client computing device.

Privacy-preserving augmented and virtual reality using homomorphic encryption

A method comprises receiving, from a remote device, an audio/video input signal, identifying one or more objects in the audio/video input signal tagged as a sensitive object, evaluating a set of workload requirements for a set of processing workloads comprising portions of the audio/video input signal, selecting one or more heavy processing workloads in the set of processing workloads to send to a compute service provider, in response to a determination that the one or more heavy processing workloads comprises one or more objects tagged as a sensitive object, encrypting the one or more objects tagged as a sensitive object using a homomorphic encryption protocol to generate a first homomorphically encrypted string, and sending the first homomorphically encrypted string to the compute service provider via a privacy protected communication channel.

On-premises augmented and virtual reality processing and privacy preserving infrastructure

One embodiment provides a parallel processor comprising a hardware scheduler to schedule pipeline commands for compute operations to one or more of multiple types of compute units, a plurality of processing resources including a first sparse compute unit configured for input at a first level of sparsity and hybrid memory circuitry including a memory controller, a memory interface, and a second sparse compute unit configured for input at a second level of sparsity that is greater than the first level of sparsity.

Programmable coarse grained and sparse matrix compute hardware with advanced scheduling

A library of machine learning primitives is provided to optimize a machine learning model to improve the efficiency of inference operations. In one embodiment a trained convolutional neural network (CNN) model is processed into a trained CNN model via pruning, convolution window optimization, and quantization.

Convolutional neural network optimization mechanism

An apparatus to facilitate optimization of a convolutional neural network (CNN) is disclosed. The apparatus includes optimization logic to receive a CNN model having a list of instructions and including pruning logic to optimize the list of instructions by eliminating branches in the list of instructions that comprise a weight value of 0.

An apparatus to facilitate workload scheduling is disclosed. The apparatus includes one or more clients, one or more processing units to processes workloads received from the one or more clients, including hardware resources and scheduling logic to schedule direct access of the hardware resources to the one or more clients to process the workloads.

Neural network scheduling mechanism

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 analyze a workload and assign the workload to one of the first type of execution unit or the second type of execution unit. Other embodiments are also disclosed and claimed.

Dynamic distributed training of machine learning models

A mechanism is described for facilitating fast data operations and for facilitating a finite state machine for machine learning at autonomous machines. A method of embodiments, as described herein, includes detecting input data to be used in computational tasks by a computation component of a processor including a graphics processor. The method may further include determining one or more frequently-used data values (FDVs) from the data, and pushing the one or more frequent data values to bypass the computational tasks.

Data operations and finite state machine for machine learning via bypass of computational tasks based on frequently-used data values

An integrated circuit (IC) package apparatus is disclosed. The IC package includes one or more processing units and a bridge, mounted below the one or more processing unit, including one or more arithmetic logic units (ALUs) to perform atomic operations.

Graphics processing integrated circuit package

Embodiments provide systems and methods which facilitate optimization of a convolutional neural network (CNN). One embodiment provides for a non-transitory machine-readable medium storing instructions that cause one or more processors to perform operations comprising processing a trained convolutional neural network (CNN) to generate a processed CNN, the trained CNN having weights in a floating-point format. Processing the trained CNN includes quantizing the weights in the floating-point format to generate weights in an integer format. Quantizing the weights includes generating a quantization table to enable non-uniform quantization of the weights and quantizing the weights from the floating-point format to the integer format using the quantization table. The operations additionally comprise performing an inference operation utilizing the processed CNN with the integer format weights.

One embodiment provides for a compute apparatus to perform machine learning operations, the compute apparatus comprising a decode unit to decode a single instruction into a decoded instruction, the decoded instruction to cause the compute apparatus to perform a complex compute operation.

One embodiment provides for a compute apparatus to perform machine learning operations, the compute apparatus comprising a decode unit to decode a single instruction into a decoded instruction, the decoded instruction to cause the compute apparatus to perform a complex machine learning compute operation.

Growing community of inventors

Rancho Cordova, CA, United States of America

Jeremy Bottleson