From 54c08c5bc3b1f7802683324f51659e71c484fbb5 Mon Sep 17 00:00:00 2001 From: Andreas Tsouchlos Date: Thu, 5 Feb 2026 03:26:54 +0100 Subject: [PATCH] Finish performance evaluation and conclusion slides --- src/midterm_presentation/main.tex | 211 ++++++++++++++++++++++++++---- 1 file changed, 185 insertions(+), 26 deletions(-) diff --git a/src/midterm_presentation/main.tex b/src/midterm_presentation/main.tex index 587a7c7..13c7274 100644 --- a/src/midterm_presentation/main.tex +++ b/src/midterm_presentation/main.tex @@ -156,6 +156,11 @@ long=spatially-coupled low-density parity-check } +\DeclareAcronym{ler}{ + short=LER, + long=logical error rate, +} + % % % Document body @@ -1737,7 +1742,7 @@ \begin{frame} \frametitle{Future Work} - \vspace*{-12mm} + \vspace*{-15mm} \begin{itemize} \item Completed work @@ -1745,6 +1750,7 @@ \item Review literature on fault-tolerant \ac{qec} using \acp{dem} \item Identify research gap + \item Familiarize with software toolboxes \end{itemize} \vspace*{7mm} \item Research gap @@ -1795,24 +1801,158 @@ \label{subsec:The Code and Other Parameters} \begin{frame} - \frametitle{Noise Models and Figures of Merit} + \frametitle{Performance Evaluation} - \begin{itemize} - \item \red{For circuit-level noise, often, all error probabilities - are set to the same value for simulations - \citereferencemanual{FSG09}} - \item \red{There are other approaches (e.g., SDMB noise, SI noise) - \citereferencemanual{DTB$^+$25}} - \end{itemize} + % - Gong et al. don't actually analyze the latency -> + % Benchmarking against other methods would be interesting - \vspace*{10mm} + % \item For circuit-level noise, use same + % \schlagwort{physical error rate} for all error + % locations \citereferencemanual{FSG09} - \begin{itemize} - \item \red{Footprint plots} - \item \red{Other figure of merit (Look into ECCentric?)} - \end{itemize} + \vspace*{-15mm} - \vspace*{15mm} + \begin{minipage}{0.35\textwidth} + \only<1>{ + \begin{itemize} + \item Independent variables + \begin{itemize} + \item Physical error rate + \item CNOT infidelity + \item Total qubit count + \item \ldots + \end{itemize} + \end{itemize} + } + \only<2->{ + \begin{itemize} + \item Independent variables + \begin{itemize} + \item \textbf{Physical error rate} + \item CNOT infidelity + \item \textbf{Total qubit count} + \item \ldots + \end{itemize} + \end{itemize} + } + \end{minipage}% + \begin{minipage}{0.65\textwidth} + \begin{itemize} + \only<1>{ + \item Noise models + \begin{itemize} + \item Standard circuit-based depolarizing noise + \citereferencemanual{FSG09} + \item Superconductor inspired (SI1000) + \citereferencemanual{GNF$^+$21} + \item Entangling Measurements (EM3) + \citereferencemanual{GNF$^+$21} + \item \ldots + \end{itemize} + } + \only<2->{ + \item Noise models + \begin{itemize} + \item \textbf{Standard circuit-based depolarizing noise} + \citereferencemanual{FSG09} + \item Superconductor inspired (SI1000) + \citereferencemanual{GNF$^+$21} + \item Entangling Measurements (EM3) + \citereferencemanual{GNF$^+$21} + \item \ldots + \end{itemize} + } + \end{itemize} + \end{minipage} + + \vspace{5mm} + + \visible<3->{ + \begin{itemize} + \item Degeneracy, information stored in correlations + $\rightarrow$ Consider \schlagwort{\acl{ler}} (LER) + \end{itemize} + } + \visible<4->{ + \begin{itemize} + \item Types of benchmarking plots + \end{itemize} + + \vspace*{5mm} + + \begin{figure}[H] + \centering + \begin{subfigure}{0.35\textwidth} + \centering + \begin{tikzpicture} + \begin{axis}[ + domain=-5:5, + width=7cm, + height=5.5cm, + xticklabels=\empty, + yticklabels=\empty, + xlabel={Physical error rate}, + xlabel style={yshift=5mm}, + ylabel={LER}, + ylabel style={yshift=-5mm}, + grid, + ] + + \addplot+[ + mark=none, + kit-red, + line width=2pt, + ] + table[row sep=crcr] { + x y \\ + 1.134800559068837 0.5575221183357257 \\ + 2.0632737437615223 0.9764009116710485 \\ + 2.861072612292603 1.7787608707489788 \\ + 3.7551580964997053 2.8407080379684153 \\ + 4.264098875196703 3.513274267363004 \\ + 4.573589936760932 3.9911505302955272 \\ + 4.903713970055305 4.268436552233389 \\ + }; + \end{axis} + \end{tikzpicture} + \end{subfigure}% + \begin{subfigure}{0.35\textwidth} + \centering + \begin{tikzpicture} + \begin{axis}[ + domain=-5:5, + width=7cm, + height=5.5cm, + xticklabels=\empty, + yticklabels=\empty, + xlabel={Total qubit count}, + xlabel style={yshift=5mm}, + ylabel={LER}, + ylabel style={yshift=-5mm}, + grid, + ] + + \addplot+[ + mark=none, + kit-blue, + line width=2pt, + ] + table[row sep=crcr] { + x y \\ + 1.147643096789246 3.8430493581808607 \\ + 1.7245658892318043 2.762331811591747 \\ + 2.3573205843145306 2.3587443650766753 \\ + 2.9156332708646624 1.560537992857378 \\ + 3.6352360073136527 1.0403588210329737 \\ + 4.392060012189421 0.7130042787942606 \\ + }; + \end{axis} + \end{tikzpicture} + \end{subfigure} + \end{figure} + } + + \vspace*{4mm} \addreferencesmanual {FSG09}{ @@ -1820,9 +1960,9 @@ ``High-threshold universal quantum computation on the surface code,'' \emph{Physical Review}, 2009. } - {DTB$^+$25}{ - P.- J. H. S. Derks et al., ``Designing fault-tolerant - circuits using detector error models,'' \emph{Quantum}, 2025. + {GNF$^+$21}{ + C. Gidney et al., ``A Fault-Tolerant Honeycomb Memory'', + \emph{Quantum}, 2021. } \stopreferencesmanual \end{frame} @@ -1838,12 +1978,27 @@ \begin{minipage}[c]{0.65\textwidth} \begin{itemize} - \item \red{Noise model} - \item \red{Memory or stability experiment} - \item \red{Figure of merit: Footprint plot} - \item \red{Comparison with BB code also simulated by - \citereferencemanual{GCR24}} - \item \red{Comparison with surface code} + \item Problem setting + \begin{itemize} + \item Research area: Decoder design for \acp{dem} + under circuit-level noise + \item Research gap: Consideration of \acp{dem} as + \ac{scldpc} codes + \end{itemize} + \vspace*{5mm} + \item Future work + \begin{itemize} + \item Modify existing decoder to pass soft information + \item Test different \ac{bp} variations + \item \ldots + \end{itemize} + \vspace*{5mm} + \item Parameters + \begin{itemize} + \item Use standard depolarizing noise for comparability + \item Compare performance with other \ac{bb} and + surface code decoders + \end{itemize} \end{itemize} \end{minipage}% \begin{minipage}[c]{0.35\textwidth} @@ -1852,9 +2007,13 @@ \begin{figure}[H] \centering - \begin{tikzpicture}[every node/.style={scale=10}] - \node at (0, 0) + \vspace*{-15mm} + \begin{tikzpicture} + \node[scale=10] at (0, 0) {\textcolor{kit-blue}{{\fontfamily{phv}\selectfont ?}}}; + + \node[align=center] at (0,-5) {Thank you for your + attention! \\ Any questions?}; \end{tikzpicture} \end{figure} \end{minipage}