Jiaming Xu Email & Phone Number

Boston, Ma, Us

Wilmington, Ma, Usa

My primary role at Onto is to maintain and continue the development of software for the JetStep series of lithography machines, particularly the newest model, the X500. The JetStep X500 has many different subsystems running on various operating systems, such as Linux, Windows, VxWorks, and Rterm. All subsystems are embedded and communicate with each other using TCP/IP sockets. My responsibilities include low-level tasks such as working on drivers and compilers, high-level tasks such as calibrations and analyses, and system-level tasks like logging and message passing between subsystems. As a result, I have strong knowledge of architecture, development, and debugging at various levels, as well as excellent communication skills and problem-solving techniques for collaborating with electronics, mechanical, and optical teams.Technologies used: Linux, VxWorks, Nvidia SDK, ClearCase, VS code, C++, C, Matlab, Python(1) Led the embedded development of real-time calibrations for the lithography machine. Designed and implemented algorithms for determining the best chuck position, best focus, and stage tip and tilt using real-time image capturing.(2) Led the maintenance of Matlab and Python scripts in the codebase. Most of the calibrations and analyses are written in the form of scripts due to the available toolboxes. I am responsible for managing these scripts and their engines.(3) Led the renovation of the remote calibration GUI, including high-speed signal plotting.(4) Contributed to the software development and testing of the next generation production line: JetStep X500 HRP.

Natick, Ma, Usa

My primary role at MathWorks was to maintain and continue the development of HDL Coder. HDL Coder is a product that can automatically translate Simulink models into HDL (Hardware Description Language, an assembly language for chip) code. HDL code is a low-level language widely used in FPGA and ASIC chips. Therefore, this product can generate thousands of lines of code from human-readable models without any errors, making it an essential tool for industrial mass production and control. I focused on optimizing the translation because the resources of the target chips are very limited, and we aimed for the highest possible performance. Two main languages are used during development: C++ for implementing functionality, and Matlab for GUI design. Development occurs on a Linux system, while testing and debugging are carried out on various common operating systems. Verilog and VHDL are the target low-level languages, so I also need to be very familiar with them to instruct HDL Coder in translation.Technologies used: Linux, Perforce, Git, VS code, C++, Matlab, Verilog, VHDL(1) Maintained the HDL Coder product and actively resolved issues, ensuring a consistently low bug count. (2) Led the reconstruction of timing controllers and digital signal filters. (3) Led the development of a script for highlighting dead code elimination. (4) Led the development of a delay balancing algorithm for bus structures. (5) Contributed to company-wide infrastructure renovations, including the transition from UDD to MCOS classes and from Perforce to Git. (6) Contributed to cluster building and testing.

Austin, Tx, Usa

Austin, Tx, Usa

My primary role at MRC was to collaborate with the material and fabrication teams to optimize the internal and external quantum efficiencies of various quantum devices. Finding the optimal combination of material parameters and device sizes was challenging due to the high variable dimension and complex quantum interaction behaviors. To address this challenge, we employed different machine learning methods to classify and predict the possible optimal ranges for the simulation software to explore. I was responsible for creating and training the models, as well as developing simple program interfaces for other team members to use and verify. With the assistance of the supercomputer at UT Austin, we achieved remarkable results, breaking many records and publishing numerous high-level papers.Technologies used: Python, C/C++, Pytorch, TensorFlow, Neural Networks, CST, Comsol, Simulated annealing, SVM(1) Led the design and characterization of a Purcell-enhanced mid-infrared light-emitting diode. (2) Collaborated with Sandia National Lab on the development of an all-dielectric nanoresonator metasurface.(3) Collaborated with UCSB's material department on research related to Gallium Nitride (GaN)-based inter-subband transition devices.

Shanghai, China

My primary role at IdeaOptics was to develop drivers and software for the spectrometer for customers and create an online spectrometer testing platform for internal users. For the former, we primarily used C++ to achieve high performance, while for the latter, we utilized Python to ensure the best scalabilityTechnologies used: Window, C++, driver, Python, Qt, Flask(1) Led the development of an online spectrometer testing platform designed to compare devices from different companies. (2) Contributed to the development of new drivers and software for advanced spectrometers.

Doctor Of Philosophy - Phd, Electrical And Computer Engineering, Gpa 3.67

25 papers, 6 patents, 13 rewards, 260+ citations,

Master Of Engineering - Meng, Electrical And Computer Engineering, Gpa 3.85

China National Scholarship Shanghai Outstanding Thesis

Bachelor Of Engineering - Be, Electrical And Computer Engineering, Gpa 3.98

China National Scholarship Shanghai Outstanding Graduate