thoughts
Short scientific thoughts, shared openly for discussion.
Short scientific thoughts that I find worth sharing: hypotheses, framings, and open questions that have shaped how I look at research problems.
At a given level of abstraction, whether physical, circuit, or microarchitectural, the leakage path is often unknown a priori. Rather than inferring it from a passive victim we neither control nor fully observe, we can invert the problem: fix a receiver (the probing primitive) and ask which sender communicates through it most effectively. Because the achievable rate is bounded by the mutual information, which grows with the effective SNR of the leakage, the sender maximizing the rate is the one that couples most strongly to the receiver. Its structure is thus shaped by the channel and carries the signature of the coupling mechanism. Reading this signature yields a testable hypothesis and guides the design of matched receivers, closing the loop that turns unknown leakage into understood leakage.
Eliott Quéré, Scientific thoughts, Thought #1, 2026. #thought-1 · Discuss this thought