Hardware Security

What property does an SRAM PUF exploit?

  1. ARight after power-on, each SRAM cell tends to settle to 0 or 1 depending on transistor variation
  2. BSRAM retains data permanently with power off
  3. CEvery SRAM cell always becomes 0 at power-on
Show answer and explanation

Answer: A. Right after power-on, each SRAM cell tends to settle to 0 or 1 depending on transistor variation

An SRAM cell is a pair of cross-coupled inverters, and which way it falls right after power-on is decided by slight strength differences between the transistors of the two inverters. So the pattern of initial values across cells is unique to each chip and reappears almost identically at every power-on. The advantage is that it uses SRAM most chips already have, with no special circuitry.

  • A. Correct. The pattern of initial values is the chip's fingerprint.
  • B. SRAM is volatile.
  • C. If all were 0, there would be no fingerprint.
Practice this topic Back to topic

More Hardware Security questions

  1. Why is error correction (such as helper data) combined with PUFs when deriving keys?
  2. Which describes a true random number generator (TRNG) built into a chip?
  3. What can happen if poor-quality, predictable random numbers are used to generate cryptographic keys?
  4. Which describes a hardware Trojan?
  5. What is the main reason hardware Trojans are hard to find with ordinary testing?