Nadav Rotem

USPTO Granted Patents = 15 (Patent Applications = 2)

Average Co-Inventor Count = 4.7

ph-index = 4

Forward Citations = 39(Granted Patents)

Company Filing History:

Years Active: 2019-2022

Loading Chart...

15 patents (USPTO):Sort by Forward CitationsExplore Patents

1. 11514306 - Static memory allocation in neural networks (2022-11-29),

2. 11487888 - Systems and methods for protecting neural network weights (2022-11-01),

3. 11264011 - Systems and methods for employing predication in computational models (2022-03-01),

4. 11256977 - Lowering hardware for neural networks (2022-02-22),

5. 11023240 - Systems and methods for efficient scaling of quantized integers (2021-06-01),

6. 10948966 - Systems and methods for optimizing power usage for systems within quality-of-service constraints (2021-03-16),

7. 10817783 - Systems and methods for efficiently updating neural networks (2020-10-27),

8. 10719613 - Systems and methods for protecting neural network weights (2020-07-21),

9. 10699190 - Systems and methods for efficiently updating neural networks (2020-06-30),

10. 10671147 - Dynamic power management for artificial intelligence hardware accelerators (2020-06-02),

11. 10579383 - Systems and methods for efficient scaling of quantized integers (2020-03-03),

12. 10553207 - Systems and methods for employing predication in computational models (2020-02-04),

13. 10482156 - Sparsity-aware hardware accelerators (2019-11-19),

14. 10474430 - Mixed-precision processing elements, systems, and methods for computational models (2019-11-12),

15. 10372787 - Hardware accelerator pre-configured with coefficients for matrix-transform operations (2019-08-06),

Title: Innovations of Nadav Rotem

Introduction

Nadav Rotem is a prominent inventor based in Santa Clara, CA, known for his significant contributions to the field of neural networks. With a total of 15 patents to his name, he has made remarkable advancements in computer science and artificial intelligence.

Latest Patents

One of his latest patents is titled "Static memory allocation in neural networks." This computer-implemented method involves compiling a neural network by organizing an interconnected set of nodes in layers. For each node, it assigns an associated activation from a plurality of activations, which outputs respective tensors. The method also allocates memory for these activations by determining the memory size for each activation and assigning memory blocks accordingly. After allocating memory, the method accesses these blocks to perform activations and execute the neural network.

Another notable patent is "Systems and methods for protecting neural network weights." This method identifies a neural network with interconnected nodes organized in layers represented by matrices containing weights. It encrypts these weights using an encryption cipher and decrypts them upon detecting the initiation of the neural network's execution.

Career Highlights

Nadav has worked with leading technology companies, including Facebook, Inc. and Meta Platforms, Inc. His experience in these organizations has allowed him to develop innovative solutions that enhance the functionality and security of neural networks.

Collaborations

Throughout his career, Nadav has collaborated with talented individuals such as Abdulkadir Utku Diril and Mikhail Smelyanskiy. These collaborations have contributed to the advancement of his research and the successful development of his patents.

Conclusion

Nadav Rotem's work in the field of neural networks showcases his innovative spirit and dedication to advancing technology. His patents reflect a deep understanding of computer science and a commitment to improving the efficiency and security of artificial intelligence systems.

This text is generated by artificial intelligence and may not be accurate.

Please report any incorrect information to support@idiyas.com