Pratheek B

Pratheek is a Senior Software Engineer at NVIDIA, working on improving GPU memory management. He obtained his Ph.D. from the Indian Institute of Science, Bengaluru in 2026. His research revolves around improving GPU memory management, focusing on address translation and data placement for large-memory workloads. His work spans the domains of GPU micro-architecture, compiler techniques, and operating systems. Previously, he had worked on reverse-engineering Nvidia GPUs. He also maintains an active interest in CPU micro-architecture research.

• SuperUVM: Enhancing GPU memory oversubscription with observability (ISCA’26)
  Pratheek B., Khushit Shah, Arkaprava Basu
• SUV: Static analysis guided Unified Virtual Memory (MICRO’24)
  Pratheek B., Guilherme Cox, Jan Vesely, Arkaprava Basu
• Designing Virtual Memory System of MCM GPUs (MICRO’22)
  Pratheek B.*, Neha Jawalkar*, Arkaprava Basu [*Equal Contribution]
• Improving GPU Multi-tenancy with Page Walk Stealing (HPCA’21) [Best paper nominee]
  Pratheek B.*, Neha Jawalkar*, Arkaprava Basu [*Equal Contribution]
• (Mis)managed: A Novel TLB-based Covert Channel on GPUs (AsiaCCS’21)
  Ajay Nayak, Pratheek B., Vinod Ganapathy, Arkaprava Basu

• Intel Ph.D. Fellowship, 2023–24
• GATE 2017: All India Rank 10/100,000
• Best Paper Nominee: HPCA 2021

• PC member: MICRO 2026
• Shadow Program Committee: EuroSys 2025
• Sub-reviewer: HPCA 2025
• Sub-reviewer: TDSC 2022

• Improving GPU memory management (Intel, Bengaluru)
• Reverse engineering NVIDIA GPU TLBs (IIT Bombay)
• Improving UVM with static analysis (IIT Gandhinagar)
• Designing virtual memory for MCM GPUs (SysTalks@India)

• Improving UVM policies under GPU memory oversubscription
• Study on impact of multi-tenancy on UVM-based GPU workloads
• Analysing replacement policies for DRAM Caches
• Utilities for porting TCP Evaluation Suite to NS3

• Email (IISc): pratheekb@iisc.ac.in
• Email (Personal): pratheekbksd@gmail.com
• Website: pratheekb.github.io
• CV: download