Analysis of Thermal Side-Channel Attacks on Analog/Digital Computing-in-Memory Accelerators
IEEE Latin American Test Symposium (LATS), pp. 1–4
Abstract
Compute-in-memory (CIM) promises to tackle the energy-efficiency barriers encountered by Von-Neuman computing, especially for machine learning workloads. One option of implementing CIM works by storing neural network weights in emerging memristor devices, connected together to form crossbars. This structure can conduct matrix-vector multiplications (MVM) in-place. The calculation causes heat to be emitted from the memristors, leading to a thermal side-channel. An attacker could sense the temperatures of the memristors, allowing recovery of the stored weights. We develop a fast and accurate thermal simulator, and evaluate the effectiveness of potential thermal side-channels. We evaluate differential weight mapping with different quantization levels. Results include recovery of weights with 70-100% accuracy for neural network layers.
Authors 3
-
Affiliation as printed
RWTH Aachen University,Institute for Communication Technologies and Embedded Systems - ICE,Germany
Institute for Communication Technologies and Embedded Systems - ICE, RWTH Aachen University, Germany
-
Affiliation as printed
RWTH Aachen University,Institute for Communication Technologies and Embedded Systems - ICE,Germany
Institute for Communication Technologies and Embedded Systems - ICE, RWTH Aachen University, Germany
-
Affiliation as printed
RWTH Aachen University,Institute for Communication Technologies and Embedded Systems - ICE,Germany
Institute for Communication Technologies and Embedded Systems - ICE, RWTH Aachen University, Germany
Cited by 1 stored of 1
1 result
No patents citing this paper on Lens.org (checked 2026-10-06).
References 18
-
W3208270918details pending0citations
-
W6766787143details pending0citations
18 results