This inventor holds 3 USPTO granted patents. Top assignee: Amazon Technologies, Inc.. Active years: 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: 2026
Title: The Innovations of Donald John Kretsch
Introduction
Donald John Kretsch is an accomplished inventor based in Cupertino, CA. He is known for his contributions to the field of integrated circuit design, particularly in the area of matrix multiplication packing. His innovative techniques have the potential to enhance the efficiency of neural network models.
Latest Patents
Kretsch holds a patent titled "Tile assignment for matrix multiplication packing." This patent describes a technique for arranging matrix multiplications for concurrent execution in an integrated circuit device. The process involves obtaining a representation of a data dependency graph of a neural network model, which includes an accumulation group pack of accumulation groups. Each of these groups contains one or more matrix multipartition instructions. The invention also details the generation of a memory location base partition constraint graph and an AG row group constraint graph, which are essential for assigning the AGs to tiles in an integrated circuit device.
Career Highlights
Donald John Kretsch is currently employed at Amazon Technologies, Inc. His work focuses on advancing technologies that improve computational efficiency and performance. His innovative approach to matrix multiplication has garnered attention in the tech community.
Collaborations
Kretsch has collaborated with notable colleagues, including Jiading Gai and Tobias Joseph Kastulus Edler Von Koch. These partnerships have contributed to the development of cutting-edge technologies in the field of integrated circuits.
Conclusion
Donald John Kretsch's contributions to the field of integrated circuit design and his innovative patent demonstrate his commitment to advancing technology. His work continues to influence the efficiency of neural network models and integrated circuit devices.
Data source: USPTO (United States Patent and Trademark Office) public patent records. Weekly synchronization. How IDiyas builds this profile
