This inventor holds 1 USPTO granted patent and 2 published patent applications. Top assignees: Advanced Micro Devices Corporation, Ati Technologies Ulc. Active years: 2026.
Company Filing History:


Years Active: 2026
Title: Alessandro Pappalardo: Innovator in Neural Network Training
Introduction
Alessandro Pappalardo is a notable inventor based in Milan, Italy. He has made significant contributions to the field of neural networks, particularly in the area of quantization-aware training. His innovative approach aims to enhance the efficiency of neural network nodes while minimizing computational intensity.
Latest Patents
Pappalardo holds a patent titled "Quantization-aware training with numerical overflow avoidance for neural networks." This invention describes an apparatus and method for creating less computationally intensive nodes for a neural network. The computing system involved includes a memory that stores multiple input data values for training the neural network, along with a processor. Instead of determining a bit width P of an integer accumulator based on the input data values and corresponding weight values, the processor selects the bit width P during the training process. It also adjusts the magnitudes of the weight values during iterative training stages to ensure that the L1 norm value of the weight values does not exceed a specified weight magnitude limit.
Career Highlights
Throughout his career, Pappalardo has worked with prominent companies in the technology sector. He has been associated with Advanced Micro Devices Corporation and ATI Technologies, where he contributed to advancements in computing technologies. His work has had a lasting impact on the development of efficient neural network systems.
Collaborations
Pappalardo has collaborated with notable professionals in his field, including Ian Charles Colbert and Mehdi Saeedi. These collaborations have further enriched his work and contributed to the advancement of neural network technologies.
Conclusion
Alessandro Pappalardo is a distinguished inventor whose work in neural network training has paved the way for more efficient computational methods. His innovative patent reflects his commitment to advancing technology in this critical area.