Bio

Dali Lai is an undergraduate student major in Electrical Engineering at Purdue University and is working toward pursuing a Ph.D. degree. He is currently exploring different kinds of fields of interests, and is most interested in VLSI and microelectronics for now. He came to the US after graduating high school and went to one year of an ESL program then spent two years in community college, fixing his GPA, and is now at Purdue. He is a PC/technology enthusiast and CPU overclocker and is interested in everything computer and modern-technology related, especially in the hardware aspect. During his free time, he plays competitive games seriously as well as downhill longboarding and beatboxing.

Computational Essay: Logic-“Digital System” Related

Project: The Performance War – PC Benchmarking Suite for WL environment

Goal

Determining how good a PC is in a universal score is always handy when it comes to comparing it to other machines. Nowadays, with each kind of focus on hardware design, it’s hard to say one machine will always out-perform the other just based on the model. Therefore, benchmarking, testing the machine performance, to determine how good it is in a certain kind of workload is necessary. In this project, benchmarking one’s machine using the Wolfram Language to see which machine performs the best is the main focus.

Summary of Results

After running the whole PC benchmark and finishing all test sets, the function outputs a report card that will contain points earned in each category as well as the total score, machine name and full time spent on the whole benchmarking program. One can compare this score to others and find out whose machine perform better in the Wolfram Language environment.

Future Work

Improving the algorithm of how the points are being calculated, making the GPU test set more intense so that high-end graphics card can differentiate each other and creating a process bar that can show which task the user is on will be helpful. Also, some C code for doing large integer calculation on GPU will be necessary. An OpenCL version of GPU benchmarking will be friendly for AMD GPU users. Also, being able to get necessary system information, such as GPU, CPU model, etc., and a ranking system for people to compete with others will be useful additions.