add ngisearch2023.tex for 17apr2023 presentation

diff --git a/conferences/ngisearch2023/ngisearch2023.tex b/conferences/ngisearch2023/ngisearch2023.tex
new file mode 100644 (file)
index 0000000..77ed3d0
--- /dev/null
+++ b/conferences/ngisearch2023/ngisearch2023.tex
@@ -0,0 +1,94 @@
+\documentclass[slidestop]{beamer}
+\usepackage{beamerthemesplit}
+\usepackage{graphics}
+\usepackage{pstricks}
+
+\graphicspath{{./}}
+
+\title{NGI SEARCH}
+\author{Luke Kenneth Casson Leighton}
+
+
+\begin{document}
+
+\frame{
+   \begin{center}
+    \huge{NGI SEARCH}\\
+    \vspace{32pt}
+    \Large{SVP64 Power ISA}\\
+    \Large{Vector Optimisation for Search}\\
+    \vspace{24pt}
+    \large{RED Semiconductor Ltd}\\
+    \large{VectorCamp}\\
+    \large{Vantosh}\\
+    \vspace{6pt}
+    \large{Assisted by Libre-SOC}\\
+  \end{center}
+}
+
+
+\frame{\frametitle{SVP64 Power ISA Vector Optimisation for Search}
+
+\vspace{6pt}
+
+ \begin{itemize}
+    \item Take existing search algorithms and optimise a Vector ISA \\
+       (at the hardware level) to\rincrease energy efficiency\\
+         \vspace{6pt}
+    \item Whilst normally a software developer has to make the best
+       out of the situation of being provided with a "fait-accomplit" 
+       ISA (ARM, x86, MIPS) and attempt to optimise Search
+    \r   algorithms for it, we aim to turn it around: work out what
+    \r   features of an ISA, if added,\r would make the same algorithms
+    \r   much more power-efficient.
+               \vspace{6pt}
+   \item RED:    Hardware/ISA concepts and project management\\
+   \item VectorCamp:    Algorithm analysis and concepts\\
+   \item VanTosh:    Validation, evolution and demonstration\\
+  \end{itemize}
+}
+
+
+\frame{\frametitle{Low-level Libraries: used by all software}
+\vspace{6pt}
+ \begin{itemize}
+   \item Regardless of the high-level algorithm portable
+               low-level libraries are needed. JSON parsing, UTF-8, strncpy, memcpy.
+               \vspace{10pt}
+   \item Optimising libc6 standard routines (strncpy, memcpy, isascii)
+                would represent high value-for-money if made much more efficient
+               \vspace{10pt}
+   \item strncpy already done under NLnet-funded Libre-SOC Grants: 11 Vector
+                instructions, representing a huge power-saving and high bang-per-buck
+               \vspace{10pt}
+   \item We want to see how far this technical approach can be taken.
+               \vspace{10pt}
+  \end{itemize}
+}
+
+\frame{\frametitle{HyperScan}
+\vspace{6pt}
+ \begin{itemize}
+   \item Hyperscan is a high-performance multiple regex match library
+               \vspace{10pt}
+  \end{itemize}
+}
+
+\frame{
+  \begin{center}
+    {\Huge The end\vspace{15pt}\\
+                  Thank you\vspace{15pt}\\
+                  Questions?\vspace{10pt}
+       }
+  \end{center}
+  
+  \begin{itemize}
+       \item Discussion: http://lists.libre-soc.org
+       \item Libera IRC \#libre-soc
+       \item https://libre-soc.org/
+       \item https://www.ngi.eu/ngi-projects/ngi-search/
+  \end{itemize}
+}
+
+
+\end{document}