Ali Farahani
Ali Farahani

Research Assistant

Biography

I received a B.Sc. in Computer Engineering from the University of Tehran in 2020 and an M.Sc. in the same major from Sharif University of Technology (SUT) in 2024. Currently, I’m a research assistant at SUT under the supervision of Prof. Hossein Asadi in the Data Storage, Networks, and Processing (DSN) lab. My research interests include Operating Systems, High-Performance Computing (HPC), Systems for Machine Learning and Distributed Systems.

Download my CV
Interests
  • Operating Systems
  • High-Performance Computing
  • Systems for ML
Education
  • MSc in Computer Engineering, 2024

    Sharif University of Technology

  • BSc in Computer Engineering, 2020

    University of Tehran

📚 My Current Research

Nowadays, machine learning is ubiquitous, from academia to industry. This miracle has revolutionized how we interact with technology and hastened the pace of innovation in everything we see around us.

Despite all advancements in recent years, the ML community is challenged with some aspects of their job that somehow slow this expeditiously growing field: I) Model training is an I/O-Intensive process, requiring multiple epochs of giant-size data loading, while the speed of I/O subsystems is far behind the other system components. II) The impacts of fine-tuning hyper-parameters in frameworks on the performance of the applications have not been sufficiently debated among ML users.

That’s where my research triggers! “Accelerating ML Workloads by Fine-Tuning In-Framework Data Loading Hyper-Parameters (Software Side) and Discovering Optimal Storage Architectures to Mitigate I/O Bottleneck (Hardware Side).”

Please reach out to collaborate if you’ve got any excellent ideas 😃

Recent News
Recent Publications
Elham Adibi, Mohammadamin Ajdari, Hossein Asadi, "Accelerating CNN Training with I/O Characterization and Fine-Tuning Data Loading Hyper-Parameters", In Preparation, 2024.