MIRACLE: MIcRo-ArChitectural Leakage Evaluation

In this paper, we describe an extensible experimental infrastructure for evaluating the micro-architectural leakage, based on power consumption, that stems from a physical device. Building existing literature, use it to systematically study 14 different devices, which span 4 instruction set architectures and vendors. The allows characterisation of each device with respect any leakage effects stemming sources within implementation. We it, example, identify document several novel (e.g., due speculative execution), scenarios where assumption about is non-portable between yet compatible devices.Ours widest its kind are aware of, highlights range challenges 1) design, implementation, evaluation e.g., masking schemes, 2) construction accurate models, 3) selection suitable devices research. For in relation 1), cast further doubt whether given upholds assumptions required by scheme; 2), conclude (statistical or formal) models must include information micro-architecture being modelled; 3), claim near mono-culture dominates literature insufficient support general claims regarding leakage. This particularly important context FIPS 140-3 standard non-invasive side-channel evaluation.