From 15ed09a2067a1a7926982cbdf499762a600b696b Mon Sep 17 00:00:00 2001 From: Alan O'Cais Date: Wed, 11 Dec 2024 16:19:45 +0100 Subject: [PATCH 1/2] Update some of the text --- isc25/EESSI/description.tex | 54 +++++++++++++++---------------------- 1 file changed, 22 insertions(+), 32 deletions(-) diff --git a/isc25/EESSI/description.tex b/isc25/EESSI/description.tex index d6cc31c..ac1180f 100644 --- a/isc25/EESSI/description.tex +++ b/isc25/EESSI/description.tex @@ -9,32 +9,22 @@ \subsection*{Overview and Goals} Application developers, HPC sites, and end users %around the world -spend significant amounts of time on optimised software installations. Surveys conducted at the -\emph{``Getting Scientific Software Installed''} Birds-of-a-Feather sessions that we (co-)organised at both SC and ISC -reveal that this (still) involves a significant amount of `manual' effort. -% (and thus time/manpower). -In the SC'19 survey, -%In the most recent survey (SC'19), -less than half of the respondents consistently automate software installation, -and only ~25\% automate environment module file generation. -%and only ~25\% automatically generate environment module files consistently, as opposed to composing them manually. -%Although the problems that arise with installing scientific software are ubiquitous, -Despite these ubiquitous problems, -there is still inadequate collaboration -between HPC sites to date: less than 30\% of respondents indicated that they -work together with other HPC sites regarding software installation, even in most recent surveys. -%Since EasyBuild can help relieve these burdens and foster collaboration, a tutorial introducing this tool is highly relevant to ISC'22 attendees. -Hence, an EESSI tutorial is very relevant to ISC'25 attendees as this tool helps relieve these burdens and fosters -collaboration. -% KH: updated to use survey results from SC19 (Tue Feb 11th 17:43 CET) -%\comment{MG: Are these numbers still OK? I seem to remember that at SC'19 more people said that they were using Spack or EB.} -%\comment{KH: 25\% Spack, 15\% EasyBuild, see slide 4 in \url{http://easybuilders.github.io/easybuild/files/SC19_GSSI_BoF/GSSI_SC19_Survey_results.pdf}; w.r.t. collaboration, it's about 33\% if you add "yes" and "sometimes" (see slide 10)} - -%The main goal of this tutorial is to facilitate the daily work of its attendees. -% It will introduce EasyBuild as a tool for providing optimised, reproducible, multi-platform scientific software installations in a consistent, efficient, and user-friendly manner. We will explain the scope of EasyBuild and show attendees how to get started, -% how to tap into some advanced features, -% showcase its use on some large-scale systems, -% and show how to engage with the EasyBuild community. +spend significant amounts of time on optimised software installations. +Much of this effort is repeatedly duplicated, not just \emph{between} +individuals but also by the individuals themselves as they reinstall +the software they need for all of the computational platforms they have access to. +%Surveys conducted at the +%\emph{``Getting Scientific Software Installed''} Birds-of-a-Feather sessions that we (co-)organised at both SC and ISC +%reveal that this (still) involves a significant amount of `manual' effort. +% In the SC'19 survey, +% less than half of the respondents consistently automate software installation, +% and only ~25\% automate environment module file generation. +% Despite these ubiquitous problems, +% there is still inadequate collaboration +% between HPC sites to date: less than 30\% of respondents indicated that they +% work together with other HPC sites regarding software installation, even in most recent surveys. +% Hence, an EESSI tutorial is very relevant to ISC'25 attendees as this tool helps relieve these burdens and fosters +% collaboration. The \textbf{European Environment for Scientific Software Installations (EESSI)}\footnote{\href{https://eessi.io}{https://eessi.io}} project is a collaborative project @@ -62,18 +52,18 @@ \subsection*{Overview and Goals} Additional families of general-purpose microprocessors including Arm 64-bit (aarch64) and RISC-V on top of th well-established Intel and AMD processors (both x86\_64), and different types of GPUS (NVIDIA, AMD, Intel) are increasing the diversity in system architectures. The rapid expansion of computational science beyond traditional -domains like physics and computational chemistry, including bioinformatis, Machine Learning (ML) and Artificial +domains like physics and computational chemistry, including bioinformatics, Machine Learning (ML) and Artificial Intelligence (AI), etc. leads to a significant growth of the software stack that is used for running scientific workloads. The emergence of commercial cloud infrastructure (Amazon EC2, Microsoft Azure, ...) and private cloud infrastructure (OpenStack) has competitive advantages over on-premise infrastructure for computational workloads, such as near-instant availability, increased flexibility, a broader variety of hardware platforms, and faster access to new -generations of microprocessors. In addition the manpower that is available in the HPC user support teams that are -responsible for helping scientists with running the software they require on high-end (and complex) infrastructure like +generations of microprocessors. However, the manpower that is available in the HPC user support teams that are +responsible for helping scientists with running the software they require on high-end and complex infrastructure like supercomputers (and beyond) is limited. These reasons indicate that there is a strong need for more collaboration on -building and installing scientific software to avoid duplicate work across computational scientists and HPC user support +building and installing scientific software to avoid duplicate work among computational scientists and HPC user support teams. -The main goal of EESSI is to provide a collection of scientific software installations that work across a wide range of +The main goal of EESSI is to provide a collection of scientific software \emph{installations} that work across a wide range of different platforms, including HPC clusters, cloud infrastructure, and personal workstations and laptops, without making compromises on the performance of that software. @@ -90,7 +80,7 @@ \subsection*{Overview and Goals} of client systems that use the software installations provided by EESSI. It consists of a limited set of libraries and tools that are installed in a non-standard filesystem location (a "prefix"), which were built from source for the supported CPU families using Gentoo Prefix. -\item The top layer of EESSI is called the software layer, which contains the actual scientific software applications +\item The user-facing layer of EESSI is called the software layer, which contains the actual scientific software applications and their dependencies. Building, managing, and optimising the software installations included in the software layer is done using EasyBuild, a well-established software build and installation framework for managing (scientific) software stacks on High-Performance Computing (HPC) systems. Next to installing the software itself, From c1de1e8a1903495beb66c55334f0d0826678e85c Mon Sep 17 00:00:00 2001 From: Alan O'Cais Date: Wed, 11 Dec 2024 16:21:47 +0100 Subject: [PATCH 2/2] Update abstract --- isc25/EESSI/abstract.tex | 20 +++++++++++--------- 1 file changed, 11 insertions(+), 9 deletions(-) diff --git a/isc25/EESSI/abstract.tex b/isc25/EESSI/abstract.tex index b342a08..5beb4b3 100644 --- a/isc25/EESSI/abstract.tex +++ b/isc25/EESSI/abstract.tex @@ -1,19 +1,21 @@ What if there was a way to avoid having to install a broad range of scientific software from scratch on every HPC cluster or cloud instance you use or maintain, without compromising on performance? -Installing scientific software for supercomputers is known to be a tedious and time-consuming task. Especially as the -HPC user community becomes more diverse, computational science expands rapidly, the diversity of system architectures -increases the application software stack continues to deepen. Simultaneously, we see a surge in interest in cloud -computing for scientific computing. Delivering optimised software installations and providing access to these +Installing scientific software for supercomputers is known to be a tedious and time-consuming task. The application +software stack continues to deepen as the +HPC user community becomes more diverse, computational science expands rapidly, and the diversity of system architectures +increases. Simultaneously, we see a surge in interest in public cloud +infrastructures for scientific computing. Delivering optimised software installations and providing access to these installations in a reliable, user-friendly, and reproducible way is a highly non-trivial task that affects application developers, HPC user support teams, and the users themselves. -This tutorial aims to address these challenges by providing the attendees with the knowledge to stream optimised -scientific software. For this, the tutorial introduces European Environment for Scientific Software Installations +This tutorial aims to address these challenges by providing the attendees with the tools to \emph{stream} the optimised +scientific software they need. The tutorial introduces European Environment for Scientific Software Installations (\emph{EESSI}), a collaboration between various European HPC sites \& industry partners, with the common goal of -creating a shared repository of scientific software installations that can be used on a variety of systems, regardless -of which flavor/version of Linux distribution or processor architecture is used, or whether it’s a full size HPC +creating a shared repository of scientific software installations (\emph{not} recipes) that can be used on a variety of +systems, regardless +of which flavor/version of Linux distribution or processor architecture is used, or whether it's a full size HPC cluster, a cloud environment or a personal workstation. We cover the usage of EESSI, different ways to accessing EESSI, how to add software to EESSI, and highlight some more -advanced features. We will also show how to engage with the community and contribute to the project. +advanced features. We will also show attendees how to engage with the community and contribute to the project.