公开(公告)号：US11244242B2

公开(公告)日：2022-02-08

申请号：US16235682

申请日：2018-12-28

Applicant: Intel Corporation

Inventor： Saurav Prakash ,  Sagar Dhakal ,  Yair Yona ,  Nageen Himayat ,  Shilpa Talwar

IPC: G06N20/00 ,  H04L12/24 ,  G06K9/62 ,  G06F9/50 ,  G06N3/08 ,  H04L12/26 ,  G06K9/00 ,  G06N7/00 ,  G06N7/08

Abstract: Systems, apparatuses, methods, and computer-readable media, are provided for distributed machine learning (ML) training using heterogeneous compute nodes in a heterogeneous computing environment, where the heterogeneous compute nodes are connected to a master node via respective wireless links. ML computations are performed by individual heterogeneous compute nodes on respective training datasets, and a master combines the outputs of the ML computations obtained from individual heterogeneous compute nodes. The ML computations are balanced across the heterogeneous compute nodes based on knowledge of network conditions and operational constraints experienced by the heterogeneous compute nodes. Other embodiments may be described and/or claimed.

公开(公告)号：US20250007600A1

公开(公告)日：2025-01-02

申请号：US18344191

申请日：2023-06-29

Applicant: Intel Corporation

Inventor： Yang-Seok Choi ,  Sagar Dhakal ,  Husam Elfadil ,  Thushara Hewavithana ,  Xiaofeng Li ,  Peng Lu ,  Tariq Qureshi ,  Jan Schreck

IPC: H04B7/08 ,  H04B7/0426 ,  H04B7/06

Abstract: Techniques are disclosed to address issues related to the computation of channel state information (CSI) and angular spectrum (AS) to perform beamforming. The CSI and AS, as well as various statistical channel parameters of a wireless channel, may be computed using different techniques, which include the use of domain knowledge enhanced neural networks (DKE-NNs). The CSI and AS may be further utilized to perform beamforming using various techniques. One of these techniques may include the implementation of eigen beamforming, which provides artificially generated power at locations within the AS that are identified with estimated eigenvector beam locations. As a result of the artificially-generated power, the resulting vector decomposition used to provide the beamforming weights results in widened eigenvector beams.

公开(公告)号：US20230068386A1

公开(公告)日：2023-03-02

申请号：US17790950

申请日：2020-12-26

Applicant: Intel Corporation

Inventor： Mustafa Riza Akdeniz ,  Arjun Anand ,  Nageen Himayat ,  Amir S. Avestimehr ,  Ravikumar Balakrishnan ,  Prashant Bhardwaj ,  Jeongsik Choi ,  Yang-Seok Choi ,  Sagar Dhakal ,  Brandon Gary Edwards ,  Saurav Prakash ,  Amit Solomon ,  Shilpa Talwar ,  Yair Eliyahu Yona

IPC: G06N20/00

Abstract: The apparatus of an edge computing node, a system, a method and a machine-readable medium. The apparatus includes a processor to perform rounds of federated machine learning training including: processing client reports from a plurality of clients of the edge computing network; selecting a candidate set of clients from the plurality of clients for an epoch of the federated machine learning training; causing a global model to be sent to the candidate set of clients; and performing the federated machine learning training on the candidate set of clients. The processor may perform rounds of federated machine learning training including: obtaining coded training data from each of the selected clients; and performing machine learning training on the coded training data.

公开(公告)号：US20190220703A1

公开(公告)日：2019-07-18

申请号：US16368716

申请日：2019-03-28

Applicant: Intel Corporation

Inventor： Saurav Prakash ,  Sagar Dhakal ,  Yair Yona ,  Nageen Himayat ,  Shilpa Talwar

IPC: G06K9/62 ,  G06N3/08 ,  G06N20/00 ,  G06F9/50

CPC classification number: G06K9/6256 ,  G06F9/5072 ,  G06N3/08 ,  G06N20/00

Abstract: Systems, apparatuses, methods, and computer-readable media are provided for load partitioning in distributed machine learning (ML) training using heterogeneous compute nodes in a heterogeneous computing environment, where the heterogeneous compute nodes are connected to a master node via respective wireless links. ML computations are performed by individual heterogeneous compute nodes on respective load partitions. The ML computations are balanced across the heterogeneous compute nodes based on knowledge of respective computational and link parameters of the heterogeneous compute nodes. Other embodiments may be described and/or claimed.