Tokey Tahmid
© Tokey © 2019 |All rights reserved

Tokey Tahmid's Homepage

About Me

Tokey Tahmid is a research associate at the Innovative Computing Laboratory (ICL) at the University of Tennessee, Knoxville. He completed his master's degree in computer science at the University of Tennessee, Knoxville. He works in the performance analysis group at ICL, where he develops Performance API (PAPI) components for specialized AI architectures such as Intel Gaudi and maintains traditional CPU and GPU components, including CUDA, ROCm, and ROCprofiler-SDK. His research focuses primarily on high-performance computing, performance engineering, and artificial intelligence.

Research

Current research work in the Performance group at ICL with Dr. Heike Jagode as the PI involves developing PAPI components for GPUs (AMD, NVIDIA, Intel) and ever-growing AI chips and accelerators such as the Intel Habana Gaudi. Research responsibilities include research work on PAPI, writing research papers, posters, and proposals.

Research work during a summer internship at the National Renewable Energy Laboratory (NREL) was on the “Low Precision and Efficient Programming Languages for Sustainable AI” role with Dr. Weslley Da Silva Pereira. The research work was published demonstrating excellent results (average speedup of 2.05X and 80.75% better energy efficiency) with mixed-precision training on multiple AI applications at NREL.

Master's thesis research was on Neuromorphic applications and Spiking Neural Networks (SNN) with Dr. Catherine Schuman as the advisor. Developed a scalable and energy-efficient infrastructure (≈22% better and ≈39% more efficient than state-of-the-art) for deep reinforcement learning (DRL) based spiking neural networks (SNN) with MPI for distributed training and mixed precision for optimization.

Research & Development Areas

  • High-Performance Computing
  • Performance Engineering
  • Performance Tools
  • AMD/ROCm Ecosystem and Tools
  • GPU Programming
  • Artificial Intelligence (AI)
  • Machine Learning & Deep Learning
  • Reinforcement Learning
  • Distributed Systems
  • Neuromorphic Computing

Publications

  • "PAPI Support for Specialized AI Architectures." – SC2025 (PDSW’25 Workshop)
  • "SpikeRL: A Scalable and Energy-efficient Framework for Deep Spiking Reinforcement Learning." – ICONS2025
  • "Energy-Efficient Computing for Scalable and Sustainable AI." – University of Tennessee 2024
  • "Towards Scalable and Efficient Spiking Reinforcement Learning for Continuous Control Tasks." – ICONS2024
  • "Low Precision for Lower Energy Consumption." – ASCR Energy Efficient Workshop 2024
  • "Low Precision and Efficient Programming Languages for Sustainable AI: Final Report for the Summer Project of 2024." – National Renewable Energy Laboratory 2024
  • "Towards the FAIR Asset Tracking Across Models, Datasets, and Performance Evaluation Scenarios." – HPEC2023
  • "Character animation using reinforcement learning and imitation learning algorithms." – ICIEV and icIVPR 2021

CV/Resume

My Resume

My Curriculum Vitae

Contact

Feel free to reach out to me via email (tokeytahmid13@gmail.com)