Updating build cache graphic and minor fix for SDK section.

Project-Management.tex

@@ -54,7 +54,7 @@ \subsubsection{The Extreme-Scale Scientific Software Stack}\label{subsubsect:e4s
 
 \begin{figure}
 	\centering
-	\includegraphics[width=0.9\linewidth]{projects/2.3.5-Ecosystem/2.3.5.01-Ecosystem-SDK/E4S_buildcache_Oct21}
+	\includegraphics[width=0.9\linewidth]{projects/2.3.5-Ecosystem/2.3.5.01-Ecosystem-SDK/E4S_buildcache_Feb22}
 	\caption{Using Spack build cache features, E4S builds can be accelerated via cached binaries for any build signature that Spack has already seen. Between September 2020 and September 2021, more than 31,000 binaries were added to the cache.}
 	\label{fig:e4s-build-cache}
 	\todo[inline]{Please add a ref to this figure in the narrative.}

projects/2.3.5-Ecosystem/2.3.5.01-Ecosystem-SDK/2.3.5.01-Ecosystem-SDK.tex

@@ -35,7 +35,7 @@ \subsubsection{\stid{5.01} Software Development Kits} \label{subsubsect:ecosyste
 
 E4S release 1.0 was announced in November 2019 on the external E4S website~\cite{e4s:homepage}. The release supports 50 ST products under Linux x86\_64. In February 2020, E4S release 1.1 extended support to both NVIDIA and AMD GPUs with the inclusion of CUDA and ROCm in a single image under Linux x86\_64. Release 1.1 also introduced support for the Linux ppc64le platform that supports CUDA 10.1. E4S releases contain HPC as well as AI/ML software including TensorFlow and PyTorch. The E4S DocPortal, accessible from the E4S website, was created to rake information from E4S product GitHub pages and provide it in a single location with the most up-to-date information about releases, installation instructions, etc. The E4S validation testsuite~\cite{e4s:validation} was introduced with support for LLVM and other ST products.
 
-In August 2021, E4S v21.11 included x86\_64 and ppc64le Docker and Singularity images with 91 E4S products. It included the DOE fork of LLVM compilers as well as E4S products built using these compilers. E4S images with support for three GPU architectures (Intel, AMD, and NVIDIA) are now available for download on DockerHub under the ecpe4s area and are released on the E4S website. These images support Intel oneAPI, AMD ROCm, and NVIDIA NVHPC and CUDA. The E4S Spack Build Cache now includes binaries for ppc64le as well as x86\_64 and includes over 77,000 total binaries (see Figure~\ref{fig:e4s-build-cache}). E4S containers now support custom images for ECP applications such as WDMapp (see Figure~\ref{fig:SpackBuildCacheWDMapp}), Nalu-Wind (see Figure~\ref{fig:E4SImageOnAWS}), and the Pantheon project (see Figure~\ref{fig:PantheonE4S}). The WBS 2.3.6.01 LANL ATDM Software Technologies project highlights the integration of Nalu-Wind and Cinema in a curated Pantheon workflow using an E4S build cache. The E4S build cache has improved the build times for these codes significantly. 
+In November 2021, E4S v21.11 included x86\_64 and ppc64le Docker and Singularity images with 91 E4S products. It included the DOE fork of LLVM compilers as well as E4S products built using these compilers. E4S images with support for three GPU architectures (Intel, AMD, and NVIDIA) are now available for download on DockerHub under the ecpe4s area and are released on the E4S website. These images support Intel oneAPI, AMD ROCm, and NVIDIA NVHPC and CUDA. The E4S Spack Build Cache now includes binaries for ppc64le as well as x86\_64 and includes over 77,000 total binaries (see Figure~\ref{fig:e4s-build-cache}). E4S containers now support custom images for ECP applications such as WDMapp (see Figure~\ref{fig:SpackBuildCacheWDMapp}), Nalu-Wind (see Figure~\ref{fig:E4SImageOnAWS}), and the Pantheon project (see Figure~\ref{fig:PantheonE4S}). The WBS 2.3.6.01 LANL ATDM Software Technologies project highlights the integration of Nalu-Wind and Cinema in a curated Pantheon workflow using an E4S build cache. The E4S build cache has improved the build times for these codes significantly. 
 \begin{figure}
         \centering
         \includegraphics[width=0.82\linewidth]{projects/2.3.5-Ecosystem/2.3.5.01-Ecosystem-SDK/E4S_WDMApp}