This inventor holds 4 USPTO granted patents and 1 published patent application. Top assignees: Neuralmagic Inc., Red Hat, Inc.. Active years: 2021-2026.
Data source: USPTO (United States Patent and Trademark Office) public patent records. Weekly synchronization. How IDiyas builds this profile
Company Filing History:
Years Active: 2021-2026
Title: Rati Gelashvili: Innovator in Neural Network Technologies
Introduction
Rati Gelashvili is a prominent inventor based in Cambridge, MA, known for his contributions to the field of neural networks. With a total of three patents to his name, Gelashvili has made significant strides in enhancing the efficiency of neural network computations.
Latest Patents
One of Gelashvili's latest patents is titled "Systems and methods for neural network convolutional layer matrix multiplication using cache memory." This invention describes a computer processor that includes multiple cores and a shared cache, optimizing the processing of neural network layers by partitioning input data into tiles and blocks. Another notable patent is "System and method of accelerating execution of a neural network." This patent outlines a method for enhancing the execution speed of neural network models by utilizing group-sparse matrices and Single Instruction Multiple Data (SIMD) tensor operations.
Career Highlights
Throughout his career, Rati Gelashvili has worked with notable companies such as NeuralMagic Inc. and Red Hat, Inc. His experience in these organizations has allowed him to develop innovative solutions that address complex challenges in neural network processing.
Collaborations
Gelashvili has collaborated with talented individuals in the field, including Alexander Matveev and Dan Alistarh. These partnerships have contributed to the advancement of his research and the successful development of his patents.
Conclusion
Rati Gelashvili's work in neural network technologies showcases his innovative spirit and dedication to improving computational efficiency. His patents reflect a deep understanding of the complexities involved in neural network processing, making him a valuable contributor to the field.
Data source: USPTO (United States Patent and Trademark Office) public patent records. Weekly synchronization. How IDiyas builds this profile
