English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Security of machine learning is increasingly becoming a major concern due tothe ubiquitous deployment of deep learning in many security-sensitive domains.Many prior studies have shown external attacks such as adversarial examplesthat tamper with the integrity of DNNs using maliciously crafted inputs.However, the security implication of internal threats (i.e., hardwarevulnerability) to DNN models has not yet been well understood. In this paper,we demonstrate the first hardware-based attack on quantized deep neuralnetworks-DeepHammer-that deterministically induces bit flips in model weightsto compromise DNN inference by exploiting the rowhammer vulnerability.DeepHammer performs aggressive bit search in the DNN model to identify the mostvulnerable weight bits that are flippable under system constraints. To triggerdeterministic bit flips across multiple pages within reasonable amount of time,we develop novel system-level techniques that enable fast deployment of victimpages, memory-efficient rowhammering and precise flipping of targeted bits.DeepHammer can deliberately degrade the inference accuracy of the victim DNNsystem to a level that is only as good as random guess, thus completelydepleting the intelligence of targeted DNN systems. We systematicallydemonstrate our attacks on real systems against 12 DNN architectures with 4different datasets and different application domains. Our evaluation shows thatDeepHammer is able to successfully tamper DNN inference behavior at run-timewithin a few minutes. We further discuss several mitigation techniques fromboth algorithm and system levels to protect DNNs against such attacks. Our workhighlights the need to incorporate security mechanisms in future deep learningsystem to enhance the robustness of DNN against hardware-based deterministicfault injections.

DeepHammer: Depleting the Intelligence of Deep Neural Networks through Targeted Chain of Bit Flips