Amir Gholaminejad

Biography

Amir Gholaminejad (Gholami), PhD, is a Research Affiliate at ICSI and an Associate Research Scientist at the Berkeley Artificial Intelligence Research and Sky Computing Labs at UC Berkeley. 

Dr. Gholami’s research focuses on large-scale machine learning and AI systems, with a focus on LLM serving (training and inference), tool usage and function calling, reasoning, and speculative methods. Previously he was part of the Nvidia team that, for the first time, made low precision neural network training possible (FP16), enabling a more than ten times increase in compute power through tensor cores. This technology has been widely adopted in today’s GPUs.

Dr. Gholami received his PhD from the University of Texas at Austin, working on large scale machine learning, a research topic for which he received UT Austin’s best doctoral dissertation award in 2018.

Publications

Selected Honors & Awards

• Melosh Medal finalist
• Amazon Machine Learning Research Award, 2020
• UT Austin’s best doctoral dissertation award, 2018
• Best student paper award, International Conference for High Performance Computing, Networking, Storage, and Analysis, 2017 
• Gold Medal in the ACM Student Research Competition
• Best student paper finalist, International Conference for High Performance Computing, Networking, Storage, and Analysis, 2014

ICSI Activity