Jason Song
Principal Systems Applications Engineer @Intersil Acquired By Renesas
Signup · Get unlimited contacts
WORK HISTORY
Principal Systems Applications Engineer @Intersil Acquired By Renesas
Tempe, AZ, US
Real-Time System Performance Classification for Power Delivery Systems using PyTorch: Developed a machine learning model using PyTorch to classify transient performance in power systems, focusing on issues such as soft-start oscillations. Implemented a neural network to analyze raw voltage and current data from hardware registers, achieving over 90% accuracy in distinguishing between good, bad, and oscillating behaviors. Achieved 99.54% precision and 98.20% recall in detecting problematic system behaviors. This system enabled real-time classification of over one million plots across varying load, temperature, and voltage conditions, significantly accelerating the hardware validation process for power delivery systems. The automated classification workflow reduced manual validation time and provided engineers with quick insights into system performance, ensuring compliance with complex power requirements.Advanced Analog Time Series Data Analysis: Achieved significant performance improvements by reducing test time by 99% in multi-phase power management systems through the automation of raw analog time series data analysis. Developed Python-based analysis pipelines and applied machine learning techniques (K-means clustering) to optimize workflows for meeting Apple\'s key power delivery requirements. Simulated specialized controller use cases for precise system performance analysis, focusing on improving power efficiency and stability in high-performance power delivery systems. This work contributed to ensuring compliance with Apple\'s stringent power management needs and enhancing system validation processes.
EDUCATION
Nanjing University
Bachelor's degree, Mathematics and Physics
Arizona State University
Master's degree, Analog and Mixed-signal Circuit Design
Georgia Institute of Technology
Master, Computer Science (specialized in Machine Learning & AI)
Nagoya University
Computer Engineering
SKILLS
ABOUT JASON SONG
Experienced Hardware/Software Principal Engineer with a strong background in machine learning, system optimization, and performance-driven semiconductor solutions. Proven ability to lead hardware validation for key power delivery systems, automate testing workflows, and develop machine learning pipelines for system-level optimizations. Expertise in integrating advanced models like LLaMA to automate decision-making processes and improve system efficiency. Proficient in Python, PyTorch, and C/C++ for high-performance, scalable designs. Adept at collaborating with cross-functional teams to deliver optimized solutions for hardware/software applications. Strong analytical skills in performance modeling and system optimization, ensuring power-efficient and high-performance hardware systems.
This profile is compiled from publicly available professional sources. Unifers is not affiliated with or endorsed by LinkedIn. Request removal of this profile.