Sergei Gofman

Title: The Innovations of Sergei Gofman

Introduction

Sergei Gofman is a notable inventor based in Haifa, Israel. He has made significant contributions to the field of deep learning and neural networks, holding a total of eight patents. His work focuses on enhancing computational models and memory access in processing systems.

Latest Patents

One of Gofman's latest patents is titled "Activation function approximation in deep neural networks using rectified-linear-unit function." This patent describes a computational apparatus designed to implement a neural network model with multiple neurons that evaluate an activation function. The apparatus includes a memory configured to hold values of a difference-function, which represent the differences between the activation function and a predefined baseline function. The circuitry evaluates the neural network model by retrieving values from memory and assessing the activation function based on these values.

Another significant patent is "Tensor-based memory access." This invention involves a processor that includes internal memory and processing circuitry. The internal memory is designed to store a definition of a multi-dimensional array located in external memory, along with indices that specify elements of the array using multi-dimensional coordinates. The processing circuitry executes instructions defined by an Instruction Set Architecture (ISA), allowing access to the multi-dimensional array through the specified coordinates.

Career Highlights

Throughout his career, Sergei Gofman has worked with prominent companies such as Habana Labs Ltd. and Intel Corporation. His experience in these organizations has contributed to his expertise in developing advanced technologies in the field of artificial intelligence and machine learning.

Collaborations

Gofman has collaborated with notable colleagues, including Shlomo Raikin and Ron Shalev. These partnerships have likely fostered innovation and the exchange of ideas in their respective projects.

Conclusion

Sergei Gofman's contributions to the field of neural networks and deep learning are significant, as evidenced by his eight patents. His work continues to influence advancements in computational models and memory access technologies.