University of California, Riverside (UCR) researchers in October conducted side-channel graphics processing unit (GPU) attacks using direct and cross-computational stacks, laying siege to multiple programs simultaneously.
Three demonstrations proved computer graphics cards could be manipulated to steal sensitive data.
UCR's Nael Abu-Ghazaleh said for one attack, the team tracked a user's Web history by checking the GPU's available memory for each page's layout and objects, with machine learning helping to identify the websites being accessed.
Abu-Ghazaleh said the other two attacks followed similar tactics, with the second tracking passwords by "observing the times of the allocations...when the user presses a key" before using "different inter-keystroke timing for different key pairs" to calculate likely combinations.
The third strategy targeted cloud networks because GPUs give users performance counters to optimize their experience; said Abu-Ghazaleh, "we recover the structure of a neural network, often a sensitive secret of companies."
From The Highlander
View Full Article
Abstracts Copyright © 2018 Information Inc., Bethesda, Maryland, USA
No entries found