This inventor holds 90 USPTO granted patents and 13 published patent applications and 3 EPO patents, primarily in Neural Network Architecture (CPC class G06N3-063). Top assignees: International Business Machines Corporation, Cornell University, Smartmotion Technologies, Inc.. Active years: 2010-2026.
Inventors with similar research interests:
Location History:
- New Windsor, NY (US) (2016)
- Mountian View, CA (US) (2018)
- San Jose, CA (US) (2021 - 2023)
- Mountain View, CA (US) (2010 - 2024)
Company Filing History:
89 out of 166,264 
2 out of 2,948 
1 out of 1 Years Active: 2010-2026
Areas of Expertise:
Title: Innovations in Neural Network Technology by John Vernon Arthur
Introduction
John Vernon Arthur, based in Mountain View, CA, is a highly accomplished inventor with an impressive portfolio consisting of 81 patents. His work primarily focuses on advancements in neural network technology, enabling significant improvements in data representation and processing capabilities.
Latest Patents
Among his most recent innovations is the patent titled "Data representation for dynamic precision in neural network cores." This invention provides a sophisticated neural network processor composed of multiple neural cores. The processor is designed to handle various processor precisions per activation, allowing it to accommodate data with different feature dimensions. Moreover, a transformation circuit is integrated into the processor, enabling the conversion of input data tensors into the appropriate precision before computation occurs across the neural cores.
Another notable patent is for "Networks for distributing parameters and data to neural network compute cores." This patent describes a neural inference chip featuring several neural cores interconnected by specialized networks. Each core is engineered to apply synaptic weights to input activations, generating output activations. The design facilitates the simultaneous delivery of weights and input data, optimizing the computational efficiency of neural networks.
Career Highlights
Throughout his career, John has worked at prestigious institutions, including IBM and Cornell University. His roles at these organizations have allowed him to contribute significantly to the field of artificial intelligence and machine learning, enhancing the effectiveness and efficiency of neural network applications.
Collaborations
John has collaborated with several prominent individuals in the field. Notable colleagues include Dharmendra S Modha and Paul A Merolla, both of whom also possess extensive expertise in the realm of neural networks and artificial intelligence.
Conclusion
John Vernon Arthur continues to be a pivotal figure in the realm of neural network technology through his innovative patents and collaboration with esteemed professionals. His inventions not only advance the field but also pave the way for future developments that can harness the potential of artificial intelligence in various applications.