{"ID":5938061,"CreatedAt":"2026-07-07T03:14:33.014478982Z","UpdatedAt":"2026-07-07T23:38:28.215472405Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2607.04055","arxiv_id":"2607.04055","title":"Securing Deep Learning Hardware: A Survey of Side-Channel Vulnerabilities and Countermeasures","abstract":"As deep learning models are increasingly deployed in critical sectors such as healthcare, finance, and security, ensuring their protection against emerging threats has become crucial. Among these threats, side-channel attacks (SCAs) represent a particular challenge since they can extract sensitive information such as model architectures, parameters, and even user inputs without requiring direct access to the model. By leveraging the physical and micro-architectural properties of the hardware, attackers can compromise systems. This survey begins by classifying leakage sources and attacker objectives, then analyzes representative studies that demonstrate practical side-channel exploits against deep-learning hardware. It also reviews existing defenses aimed at mitigating these vulnerabilities and concludes by outlining key open research challenges and potential future directions.","short_abstract":"As deep learning models are increasingly deployed in critical sectors such as healthcare, finance, and security, ensuring their protection against emerging threats has become crucial. Among these threats, side-channel attacks (SCAs) represent a particular challenge since they can extract sensitive information such as m...","url_abs":"https://arxiv.org/abs/2607.04055","url_pdf":"https://arxiv.org/pdf/2607.04055v1","authors":"[\"Zahra Mohammadi\",\"Mona Hashemi\",\"Siamak Mohammadi\"]","published":"2026-07-04T23:45:25Z","proceeding":"cs.CR","tasks":"[\"cs.CR\"]","methods":"[]","has_code":false}