Lidong Zhou

Title: Lidong Zhou - Innovator in Neural Network Technologies

Introduction: Lidong Zhou is an accomplished inventor based in Beijing, China. He is currently associated with Microsoft Technology Licensing, LLC. With a focus on advancing neural network technologies, Lidong has made significant contributions to the field through his innovative patent applications.

Latest Patent Applications: Lidong Zhou has filed several notable patent applications, including:

1. **Performing Dynamic Sparse Computation on Dense Computation-Efficient Computing Devices**: This application describes techniques for processing dynamically sparse neural networks as dense computations. It involves translating an input tensor from a sparse format into a dense format, allowing for efficient dense computation. The output data can then be converted back into the sparse format through a reverse permutation.

2. **Parallelization Plan Generation for a Neural Network**: This application outlines methods for designing and generating a parallelization plan for neural networks. It enables the distribution of workloads across multiple devices by transforming operators and tensors into functionally equivalent forms, which are then scheduled for execution on separate devices.

Conclusion: Lidong Zhou's work in the realm of neural networks showcases his innovative spirit and dedication to advancing technology. His contributions through patent applications reflect his commitment to enhancing computational efficiency in complex systems.