Company Filing History:
Years Active: 2025
Title: Pranav Jonnalagadda: Innovator in Machine Learning Technologies
Introduction
Pranav Jonnalagadda is a notable inventor based in San Jose, CA (US). He has made significant contributions to the field of machine learning, particularly in memory management and graph partitioning for machine learning models. With a total of 2 patents, his work is paving the way for advancements in efficient machine learning systems.
Latest Patents
Pranav's latest patents include a "Method and system for memory management within machine learning inference engine." This innovative method involves receiving a machine learning network model in high-level code and generating an internal representation that is mapped to components in a multi-processing tile device. The process includes determining memory allocation based on the dimensions of input/output tensor data across processing tiles, ultimately linking memory address ranges to form a grouped memory space.
Another significant patent is the "Method and apparatus for ML graphs by a compiler." This system focuses on splitting a machine learning graph, where a compiler generates an internal representation of the ML model. The graph is partitioned into subgraphs associated with different hardware, allowing for optimized execution of low-level instructions.
Career Highlights
Pranav Jonnalagadda is currently employed at Marvell Asia Pte., Ltd., where he continues to innovate in the field of machine learning. His work is instrumental in enhancing the performance and efficiency of machine learning applications.
Collaborations
Throughout his career, Pranav has collaborated with notable colleagues such as Ulf Hanebutte and Chien-Chun Chou. These collaborations have contributed to the development of cutting-edge technologies in machine learning.
Conclusion
Pranav Jonnalagadda's contributions to machine learning through his patents and work at Marvell Asia Pte., Ltd. highlight his role as a key innovator in the industry. His advancements in memory management and graph partitioning are set to influence the future of machine learning technologies.