Matthew Williams Email & Phone Number

Vancouver, British Columbia, Canada

Montreal, Quebec, Canada

I am modifying the RISC-V Instruction Set Architecture by adding custom instructions and hardware elements and altering the datapaths to enable stochastic computing directly on the chip as a part of a research project. Once the modifications have been made, I will simulate the modified RISC-V CPU on a Xilinx FPGA and benchmark its stochastic performance using Sobel Filtering processes.

Montreal, Quebec, Canada

Resolved timing violations in ASIC designs by parsing Tempus delay reports using RegExp and identifying key issues.Explored different areas of silicon design including power management, floorplanning, layout, synthesis, clock tree synthesis, power gating, clock gating and IR drop.

Montreal, Quebec, Canada

I was responsible for the modification and merging of several different firmware versions of these dedicated FPGA cards, creating one master firmware written in C++.I also had the opportunity to become familiar with industry tools and methods such as Agile development, task management with Jira, Excel, continuous integration with Jenkins, and version control with SVN.

Dorval, Quebec, Canada

Worked in the Technical Support department of Matrox alongside an OEM Engineer, primarily focusing ontheir encoding and decoding expansion card products, namely IPX, Maevex, Extio and their Display Walltechnology that uses their graphics cards and Quad-Head-2-Go products.• Created a tool in C# using .NET Framework that automatically connects to the customer’s encoder/decoderIPX controller via SSH, runs a series of tests and collects diagnostic data using Matrox’s in-house API and savesthe data in an encrypted folder on the customer’s computer.• Fixed numerous bugs in the API used by Matrox’s encoder/decoder controllers• Assisted my coworkers in the Technical Support department by helping diagnose customers’ problems

Montreal, Quebec, Canada

Our team of student researchers simulated a 2D antiferromagnetic Ising model using Python, saved its cooling history and inputted that data into a Convolutional Neural Network that we built to determine the magnitude and the shape of the magnetic field applied to Ising model.