•  About
    • About the Lab
    • Director’s Note
    • Our Vision
    • Founding Donor
    • Advisory Board
    • Principal Investigators
  • People
    • Associated Faculty
    • Executive Committee
    • Students
    • Program Directorate
  • TrustNet
  •  Projects
  •  Resources
    • Pre-Doctoral Program
    • Internships
    • Early Career Award
    • Trust Lab Grant
    • Trust Lab Fellowship
  •  News
    • Trust Matters
    • Quick Updates
  •  Events
    • Talks
    • Trust Summit
    • TL CTF
    • Schools
    • All Events
  •  Engage
  •  About
    • About the Lab
    • Director’s Note
    • Our Vision
    • Founding Donor
    • Advisory Board
    • Principal Investigators
  • People
    • Associated Faculty
    • Executive Committee
    • Students
    • Program Directorate
  • TrustNet
  •  Projects
  •  Resources
    • Pre-Doctoral Program
    • Internships
    • Early Career Award
    • Trust Lab Grant
    • Trust Lab Fellowship
  •  News
    • Trust Matters
    • Quick Updates
  •  Events
    • Talks
    • Trust Summit
    • TL CTF
    • Schools
    • All Events
  •  Engage

Friend or Foe: Belady Pursuers and Randomized Secure Last-level Caches

Overview
People
Outcome
Overview

State-of-the-art micro-architectural attacks target different on-chip and off-chip components like processor pipelines, branch predictors, caches, DRAMs, interconnects, etc. One of the solutions proposed recently is to temporally or spatially partition these resources shared among different processes and cores in the form of cache partitioning, DRAM channel partitioning, etc. However, preliminary studies show that this kind of strong isolation at all levels of microarchitecture can lead to a performance overhead in the range of 2X to 30X. Another approach to mitigate this problem is through the randomization of microarchitecture units like randomized branch predictors, caches, and TLBs that make the life of an attacker difficult. Randomization techniques incur minimal performance loss. However, these techniques incur significant storage overhead with additional design complexity.As part of the project, students will explore techniques that can provide strong security guarantees mitigating various microarchitecture attacks without compromising performance and storage overheads.
Active from 2023
Funding: Trust Lab Grant 2023

People

Biswabandan Panda

Outcome
Twitter Facebook-f Linkedin Youtube
  • trustlabcse.iitb.ac.in
  • +91-22-2159-6725
  • First Floor, New CSE Building
    Department of Computer Science and Engineering,
    Indian Institute of Technology Bombay,
    Powai, Mumbai 400076
IITB logo