Hardware Security: Advanced Topics and Trends

Which correctly explains why Rowhammer is said to become more likely as DRAM is scaled down?

  1. AScaling removes the need for refresh, so long periods pass without the cell charge ever being replenished
  2. BCells get closer together and store less charge, so they are more easily affected by interference from activations of neighbouring rows
  3. CIn scaled DRAM the standard forbids activating the same row many times, and violating that limit causes malfunctions
Show answer and explanation

Answer: B. Cells get closer together and store less charge, so they are more easily affected by interference from activations of neighbouring rows

Each time a row is activated, the word-line potential swings widely, making charge leak more easily from cells in nearby rows. As scaling shrinks the distance between cells, the interference becomes stronger, while the smaller cell capacitance also reduces the margin before a bit flips. Research has reported a tendency for newer DRAM generations to need fewer activations before the first bit flip. For this reason, countermeasures against Rowhammer become more important with each DRAM generation.

  • A. Refresh is still needed after scaling; if anything, the margin for retaining data becomes smaller.
  • B. Correct. Both the stronger interference and the smaller margin contribute.
  • C. Repeatedly activating the same row is permitted as normal operation.
Practice this topic Back to topic

Books on Hardware Security: Advanced Topics and Trends Ad

As an Amazon Associate I earn from qualifying purchases.

More Hardware Security: Advanced Topics and Trends questions

  1. Which correctly explains why Rowhammer protection by in-DRAM TRR (target row refresh) has been bypassed in research?
  2. What is wrong with the idea that "if ECC memory is used, there is no need to worry about Rowhammer"?
  3. Which correctly describes the idea of PRAC (per-row activation counting), added to the DDR5 standard updated in 2024 (JESD79-5C)?
  4. For simplicity, take the shortest interval at which rows in the same bank can be activated to be tRC = 50 ns. With a refresh period of 64 ms, the maximum number of times a neighbouring row can be activated between one refresh of a given row and the next is 64 ms ÷ 50 ns. What is this maximum number when the refresh rate is doubled?
  5. Which correctly describes the characteristics of the vulnerability called MDS (Microarchitectural Data Sampling), published in 2019?