Yahav Shadmiy

This inventor holds 12 USPTO granted patents and 4 published patent applications and 1 EPO patent. Top assignee: Intel Corporation. Active years: 2019-2026.

USPTO Granted Patents = 12 (Patent Applications = 4)

% Patents Active = 100.0

EPO Patents = 1 (Patent Applications = 5)

Average Co-Inventor Count = 11.8

ph-index = 3

Forward Citations = 16(Granted Patents)

Company Filing History:

12 out of 56,290

Intel Corporation

patents

Years Active: 2019-2026

Loading Chart...

12 patents (USPTO):Sort by Forward CitationsExplore Patents

1. 12675681 - Smart memory handling and data management for machine learning networks (2026-07-07),

2. 12131507 - Low rank matrix compression (2024-10-29),

3. 11886984 - Variable precision and mix type representation of multiple layers in a network (2024-01-30),

4. 11763140 - Smart memory handling and data management for machine learning networks (2023-09-19),

5. 11620766 - Low rank matrix compression (2023-04-04),

6. 11600035 - Sub-graph in frequency domain and dynamic selection of convolution implementation on a GPU (2023-03-07),

7. 11250610 - Sub-graph in frequency domain and dynamic selection of convolution implementation on a GPU (2022-02-15),

8. 11093822 - Variable precision and mix type representation of multiple layers in a network (2021-08-17),

9. 11087206 - Smart memory handling and data management for machine learning networks (2021-08-10),

10. 11037330 - Low rank matrix compression (2021-06-15),

11. 10762685 - Sub-graph in frequency domain and dynamic selection of convolution implementation on a GPU (2020-09-01),

12. 10467795 - Sub-graph in frequency domain and dynamic selection of convolution implementation on a GPU (2019-11-05),

Title: Innovations of Yahav Shadmiy

Introduction

Yahav Shadmiy is a prominent inventor based in Ramat Gan, Israel. He has made significant contributions to the field of technology, particularly in the area of data compression and neural networks. With a total of 11 patents to his name, Shadmiy continues to push the boundaries of innovation.

Latest Patents

One of his latest patents is focused on low rank matrix compression. This invention includes an apparatus that comprises logic, at least partially including hardware logic, to implement a lossy compression algorithm. This algorithm utilizes a data transform and quantization process to compress data in a convolutional neural network (CNN) layer. Other embodiments are also disclosed and claimed. Another notable patent is related to variable precision and mixed type representation of multiple layers in a network. This apparatus comprises a plurality of execution units, including at least a first type and a second type of execution unit. It features logic, at least partially including hardware logic, to expose embedded cast operations in at least one of a load instruction or a store instruction. The invention determines a target precision level for the cast operations and loads them at that target precision level. Other embodiments are also disclosed and claimed.

Career Highlights

Yahav Shadmiy is currently employed at Intel Corporation, where he applies his expertise in developing cutting-edge technologies. His work has significantly impacted the efficiency and performance of various computing systems.

Collaborations

Shadmiy has collaborated with notable colleagues, including Jacob Subag and Yaniv Fais, contributing to a dynamic and innovative work environment.

Conclusion

Yahav Shadmiy's contributions to technology through his patents and work at Intel Corporation highlight his role as a leading inventor in the field. His innovative approaches to data compression and neural networks continue to shape the future of technology.