Hello! I’m a PhD student in the Hardware Systems Collective at University California, Santa Cruz in Computer Science and Engineering, advised by Professor Dustin Richmond. I am interested in researching Computer In Memory Accelerators for ternary machine learning models. My current project is implementing a Scalable MatMul-Free Large Language Model accelerator using ternary weights on an FPGA and GPU. In addition, I am also doing an in depth evaluation of the performance of Ternary Machine Learning Model performance on GPUs.

I received my bachelor’s in Computer Engineering from Tufts University. During which, I was a research assistant in the Tufts Computer Architecture Lab on the HotGauge Project under Professor Mark Hempstead. My project focused on accurately simulating the impacts of diamond heat sinks on the performance of Intel x86.

The best way to contact me is through email: abban16204@gmail.com or eabban@ucsc.edu.

Research interests

• Machine Learning Acceleration
• Compute In Memory
• Reconfigurable Accelerators