Commit 4b51f85b authored by David Thompson's avatar David Thompson

A github-pages website for sensei.

parent 63f8d862
*.sw[a-z]
doc/build-tmp/
node_modules/
module.exports = {
baseUrl: '/sensei',
work: './build-tmp',
config: {
title: 'sensei',
description: '"Scalable in situ visualization and analysis."',
subtitle: '"Lightweight, zero-copy simulation adaptor."',
author: 'sensei',
timezone: 'UTC',
url: 'https://kitware.github.io/sensei',
root: '/sensei/',
github: 'kitware/sensei',
markdown: {
gfm: false,
},
},
};
---
date: 2018-05-04 16:35
markdown:
gfm: false
---
# Website
In order to make the SENSEI website more dynamic, we will start using github pages to host the website.
The website is generated from files in the `doc` directory of the sensei repository.
The process for updating the website is:
+ Ensure you have installed [nodejs][] and [npm][] and run the one-time setup command:
`npm install` from the `doc` directory of your sensei clone.
This will create a `node_modules` directory inside `doc` containing
javascript libraries needed to generate and publish the website.
You only need to do this once per clone of sensei.
+ Add new files to the `doc/content` directory as needed (more on this below).
+ Test the new website by running `npm run doc:www` and visit http://127.0.0.1:4000/
in your web browser. Verify that the page you've added or edited appears properly.
+ Commit your changes to the master branch of sense by submitting a merge request.
+ Once the changes are in master, you should _publish_ the changes to the `gh-pages`
branch of the repository. For the moment, this is a manual process: run
`npm run doc:publish` in the `doc` directory of your clone.
## Adding new files
First of all, **do not add large or binary files** to the repository.
If you want to include images, link to PDFs, etc. then upload them
to the [sensei in situ collection][] on https://data.kitware.com/
and reference them by URL in the markdown files you add here.
There are several kinds of files you might wish to add to sensei:
+ To add a blog post, simply add a new markdown file in `doc/content/_posts`.
If you put a header at the top of the file like so:
```
---
date: 2018-05-04 16:48:00
---
Markdown for page starts here
```
then the generated website will preserve the date of the original post even
if the file is later changed.
+ To add a static page to the introductory (_Learn more_) or developer (_Develop_)
sections, you will need to
+ Add a new file to the `doc/content/learn-more` or `doc/content/develop` directories.
+ Add an entries to the `doc/tpl/__en__` and `doc/tpl/__sidebar__` files that
map the filename to a title to show in the sidebar links of the website.
+ To add a new section to the links at the top of the website (a sibling to the
_Learn more_ and _Develop_ links), you'll need to do the above _plus_ add an entry
to the `doc/data/menu.yml` file.
For more information about formatting files, see the
[Markdown](https://daringfireball.net/projects/markdown/) and
[Hexo](https://hexo.io/docs/writing.html) documentation.
Markdown is the format we use for pages while Hexo is the
javascript package that kw-doc uses to process the markdown
content into HTML.
## Description of documentation files
+ `package.json` and `package-lock.json` are files that specify which javascript packages
are needed to generate and publish the documentation.
+ `doc/config.js` is javascript declaring website-wide configuration.
+ `doc/data/menu.yml` declares links at the top of each web page on the site.
+ `doc/tpl/__sidebar__` organizes the sidebar links for static sections of the website.
+ `doc/tpl/__en__` maps filenames and sidebar names into different languages/locales.
+ `doc/content/index.jade` is a template used to construct the front page of the website.
[nodejs]: https://nodejs.org/
[npm]: https://npmjs.org/
[sensei in situ collection]: https://data.kitware.com/#collection/5a007cb58d777f31ac64ddfd
# Overview
---
markdown:
gfm: false
---
# SENSEI dataset schema and ADIOS
VTK has two ways of representing parallel distributed data. In the first each
MPI rank has a single object derived from
[vtkDataSet](https://www.vtk.org/doc/nightly/html/classvtkDataSet.html). We
......
# Contributing to SENSEI
The SENSEI project uses [Kitware's GitLab][gitlab] instance to accept merge requests
rather than GitHub's pull request mechanism so that we can run automated [tests][]
on hardware with access to a graphics accelerators in a secure fashion.
[gitlab]: https://gitlab.kitware.com/sensei/sensei
[tests]: /sensei/develop/testing.html
date: 2018-05-04
marked:
gfm: false
---
# Developing with SENSEI
SENSEI is a framework that accepts data from a simulation in VTK format.
# CTest Regression Dashboard #
A number of systems have been confiugured for nightly testing and continuous integration. To view the results of these runs navigate your web browser to [http://cdash.hpcvis.com/index.php?project=SENSEI]().
# CTest Regression Dashboard
A number of systems have been confiugured for nightly testing and continuous integration.
To view the results of these runs navigate your web browser to [http://cdash.hpcvis.com/index.php?project=SENSEI]().
# Running the tests
# Running the tests #
To enable the regression tests one must configure the build with tetsing enabeld.
```bash
cmake -DBUILD_TESTING=ON ...
......@@ -16,8 +19,10 @@ To run the tests and submit the results to the web dashboard add the ctest track
ctest -DExperimental
```
# Adding regression tests using CTest #
# Adding regression tests using CTest
### senseiAddTest
Tests are added by calling the CMake function *senseiAddTest*. This function
encapsulates the common scenarios needed to compile, link, and run tests in
serial, and parallel; and configures CTest to flag absolute test failures
......@@ -56,7 +61,9 @@ senseiAddTest(testHistogramParallel
COMMAND ${MPIEXEC} ${MPIEXEC_NUMPROC_FLAG}
${MPIEXEC_MAX_NUMPROCS} testHistogram)
```
#### Python
Here is an example of a Python test.
```CMake
senseiAddTest(testADIOSFlexpath
......@@ -67,5 +74,7 @@ senseiAddTest(testADIOSFlexpath
FEATURES ${ENABLE_PYTHON} ${ENABLE_ADIOS})
```
# Setting up a new test system #
Examples of setting up nightly and continuous test sites is beyond the scope of this document. However, a few examples runs can be found at [https://github.com/burlen/SENSEI_ctest]().
# Setting up a new test system
Examples of setting up nightly and continuous test sites is beyond the scope of this document.
However, a few examples runs can be found at [https://github.com/burlen/SENSEI_ctest]().
layout: index
description: SENSEI in situ
subtitle: SENSEI ∙ Scalable in situ analysis and visualization
cmd: git clone https://gitlab.kitware.com/sensei/sensei.git
comments: false
---
ul#intro-feature-list
li.intro-feature-wrap
.intro-feature
.intro-feature-icon
i.fa.fa-cloud-download
h3.intro-feature-title
a(href="https://gitlab.kitware.com/sensei/sensei/tags").link Releases
p.intro-feature-desc SENSEI source code releases are tagged and downloadable here.
li.intro-feature-wrap
.intro-feature
.intro-feature-icon
i.fa.fa-book
h3.intro-feature-title
a(href="https://data.kitware.com/#collection/5a007cb58d777f31ac64ddfd/folder/5a049b808d777f31ac64e77d").link Tutorial
p.intro-feature-desc Tutorial material from Supercomputing 2017, including a virtual machine.
li.intro-feature-wrap
.intro-feature
.intro-feature-icon
i.fa.fa-tachometer
h3.intro-feature-title
a(href="http://cdash.hpcvis.com/index.php?project=SENSEI").link Testing dashboard
p.intro-feature-desc SENSEI is tested as commits are merged.
li.intro-feature-wrap
.intro-feature
.intro-feature-icon
i.fa.fa-bug
h3.intro-feature-title
a(href="http://gitlab.kitware.com/sensei/sensei/issues").link Issues
p.intro-feature-desc Report a bug or request a feature.
li.intro-feature-wrap
.intro-feature
.intro-feature-icon
i.fa.fa-comments-o
h3.intro-feature-title
a(href="http://discourse.kitware.com/c/sensei").link Community & Feedback
p.intro-feature-desc Tell us what you think!
div#intro-get-started-wrap
a(href="/sensei/learn-more/index.html", id="intro-get-started-link") Learn more
div#banner
img(
src="https://data.kitware.com/api/v1/file/5aea33018d777f0685797021/download"
alt="ParaView/Catalyst on 256,000 processes on Mira"
style="width: 100%;")
p.intro-feature-desc ParaView/Catalyst on 256,000 processes on Mira.
title: "About SENSEI"
---
This project takes aim at a set of research challenges for enabling scientific knowledge discovery within the context of in situ processing at extreme-scale concurrency. This work is motivated by a widening gap between FLOPs and I/O capacity which will make full-resolution, I/O-intensive post hoc analysis prohibitively expensive, if not impossible.
We focus on new algorithms for analysis, and visualization – topological, geometric, statistical analysis, flow field analysis, pattern detection and matching – suitable for use in an in situ context aimed specifically at enabling scientific knowledge discovery in several exemplar application areas of importance to DOE.
Complementary to the in situ algorithmic work, we focus on several leading in situ infrastructures, and tackle research questions germane to enabling new algorithms to run at scale across a diversity of existing in situ implementations.
Our intent is to move the field of in situ processing in a direction where it may ultimately be possible to write an algorithm once, then have it execute in one of several different in situ software implementations. The combination of algorithmic and infrastructure work is grounded in direct interactions with specific application code teams, all of which are engaged in their own R&D aimed at evolving to the exascale.
## Impact
This approach blends algorithmic R&D with focused solutions for important science problems, and pushes the limits of existing in situ infrastructure as applied to DOE science problems on extreme-concurrency platforms. This work will likely have immediate impact on several science areas through our collaborations, who are in desperate need of new analysis methods resulting for data of increasing size and complexity, as well as longer-term impact due to the emphasis on wider dissemination and distribution of these new in situ analysis algorithms as part of several different in situ frameworks. The result is an increased lifespan of software investments in key infrastructure.
## Science-Facing Projects
A primary driver for our in situ analysis algorithm R&D stems from science needs:
+ Simulations are increasingly multi-scale both in time and space;
+ The data they compute are increasingly complex;
+ It is increasingly impractical to do full-resolution I/O;
+ Data is not saved nor analyzed resulting in lost science.
To identify needs and evaluate solutions, we are using several science-facing projects, and their attendant science drivers, to shape a new generation of in situ analysis algorithm development.
## _In Situ_ Infrastructure
The in situ infrastructure thrust focuses on four in situ frameworks, namely Catalyst, Libsim, ADIOS, and GLEAN. While these infrastructures provide the means for performing various types of processing, analysis, and visualization operations from a live-running simulation, they differ in their approach to interfacing with the simulation, with how they are configured, and how they are extended by user-supplied code.
An infrastructure-based goal of the project is:
+ To achieve write-once, run-anywhere analysis.
A step to achieve this vision is by identifying the building blocks to design and refactor analysis codes on diverse in situ infrastructures as well as on a wide-variety of systems. By achieving this vision, we can enable analysis algorithms to fully exploit the underlying concurrency and heterogeneity of the supercomputing systems.
## Funding
This work is supported by the
Director, Office of Science,
Office of Advanced Scientific Computing Research,
of the U.S. Department of Energy under Contract No. DE-AC02-05CH11231,
through the grant
“Scalable Analysis Methods and In Situ Infrastructure for Extreme Scale Knowledge Discovery,”
program manager Dr. Lucy Nowell.
<!-- extra line breaks to prevent footer from obscuring text -->
<br><br><br>
# SENSEI in situ
## Overview
This project takes aim at a set of research challenges for enabling scientific knowledge discovery within the context of in situ processing at extreme-scale concurrency. This work is motivated by a widening gap between FLOPs and I/O capacity which will make full-resolution, I/O-intensive post hoc analysis prohibitively expensive, if not impossible.
We focus on new algorithms for analysis, and visualization – topological, geometric, statistical analysis, flow field analysis, pattern detection and matching – suitable for use in an in situ context aimed specifically at enabling scientific knowledge discovery in several exemplar application areas of importance to DOE.
Complementary to the in situ algorithmic work, we focus on several leading in situ infrastructures, and tackle research questions germane to enabling new algorithms to run at scale across a diversity of existing in situ implementations.
Our intent is to move the field of in situ processing in a direction where it may ultimately be possible to write an algorithm once, then have it execute in one of several different in situ software implementations. The combination of algorithmic and infrastructure work is grounded in direct interactions with specific application code teams, all of which are engaged in their own R&D aimed at evolving to the exascale.
## Funding
This work is supported by the Director, Office of Science, Office of Advanced Scientific Computing Research, of the U.S. Department of Energy under Contract No. DE-AC02-05CH11231, through the grant “Scalable Analysis Methods and In Situ Infrastructure for Extreme Scale Knowledge Discovery,” program manager Dr. Lucy Nowell.
<!-- extra line breaks to prevent footer from obscuring text -->
<br><br><br>
title: "Presentations"
last_modified_at: 2018-04-17T19:35:00-05:00
toc: true
---
# Talks
## 2016
+ Brad J. Whitlock. In Situ Production of Extract Databases for Visualization. In _ISC Workshop on In Situ Visualization_, June 2016.
+ Gunther H. Weber. IsoFind: Halo Finding Using Merge Trees. In _Department of Energy Computer Graphics Forum (DOE CGF)_, Monterey, CA, USA, May 2016.
+ Nicola J. Ferrier. Multi-modal 3D Characterization of Materials. Invited talk at the APS Horizons Workshop: Challenges in Integrating Data Science, Computational Modelling and Advanced Characterization, May 2016.
+ E. Wes Bethel. In Situ Methods and Infrastructure: Answers Without All the I/O. In _SIAM Parallel Processing 2016 (PP16)_, Paris, France, April 2016.
+ Julien Jomier and Patrick O’Leary. In Situ Methods and Infrastructure: Answers Without All the I/O. In _SIAM Parallel Processing 2016 (PP16)_, Paris, France, April 2016.
+ E. Wes Bethel. Exascale Computing Challenges and Volume Rendering Optimizations for Advanced Architectures. In _Computer Science Department Colloquium Series_, Old Dominion University, Norfolk, VA, USA, April 2016.
+ Nicola J. Ferrier. Multi-scale, multi-modal Dynamic 3D Imaging. Keynote presentation at Multimodal Data Analysis Hackathon, April 2016.
+ Brad J. Whitlock. In Situ Infrastructure Enhancements for Data Extract Generation. In _54th AIAA Aerospace Sciences Meeting_, SciTech 2016, January 2016.
## 2015
+ E. Wes Bethel. In Situ Methods: Hype or Necessity? In _IEEE International Conference for High Performance Computing, Networking, Storage and Analysis (SC15)_, Austin, TX, USA, November 2015.
+ Patrick O’Leary. In Situ Methods: Hype or Necessity? In _IEEE International Conference for High Performance Computing, Networking, Storage and Analysis (SC15)_, Austin, TX, USA, November 2015.
+ Venkatram Vishwanath. In Situ Methods: Hype or Necessity? In _IEEE International Conference for High Performance Computing, Networking, Storage and Analysis (SC15)_, Austin, TX, USA, November 2015.
+ Patrick O’Leary. Responsive large data analysis and visualization with the ParaView ecosystem. In _NVIDIA GPU Technology Theater SC15 the International Conference for High Performance Computing Networking, Storage and Analysis_, Austin, TX, November 2015.
+ Andrew Bauer. In Situ Analysis and Visualization for Rotor Aeromechanics Simulations. In _ISAV 2015: Lightning Talk_, Austin, TX, November 2015.
+ E. Wes Bethel. Visualization, Analysis, and Exascale: Trouble or Triumph? In _The Ultrascale Visualization Workshop_, IEEE International Conference for High Performance Computing, Networking, Storage and Analysis (SC15), Austin, TX, USA, November 2015.
+ Patrick O’Leary and Sebastien Jourdain. Is the web ready for visualization? In _Workshop on Visualization Technologies at SC15 the International Conference for High Performance Computing Networking, Storage and Analysis_, Austin, TX, November 2015.
+ E. Wes Bethel. HPC Visualization and Analysis at the Exascale: Big Headaches, Big Opportunities. In _SIGGRAPH Asia Symposium on Visualization in High Performance Computing_, Kobe, Japan, November 2015.
+ E. Wes Bethel. In Situ 2020: Back to the Future, Again. In _IEEE Symposium on Large Data Analysis and Visualization Symposium_, co-located with IEEE Visualization 2015, Chicago, IL, USA, October 2015.
+ Gunther H. Weber. Scientific Visualization of Big Data. In _Learning from other domains – Big Data and Visualization, 4th Workshop on Next-Generation Analytics for the Future Power Grid_, Richland, WA, USA, September 2015.
+ Earl P.N. Duque. Accelerating the Post-Processing of Large Scale Unsteady CFD Applications via In Situ Data Reduction and Extracts. In _Virginia Polytechnic University Department of Aerospace and Ocean Engineering Graduate Seminar_, Virginia Polytechnic University, Blacks- burg, VA, USA, September 2015.
+ Andrew Bauer. ParaView Catalyst: Scalable In Situ Processing. In _2015 Air Force HPC User Forum: Visualization Session_, Dayton, OH, July 2015.
+ Dmitriy Morozov. Parallel computation of persistent homology using the blowup complex. In _ACM Symposium on Parallelism in Algorithms and Architectures (SPAA)_, Portland, OR, June 2015.
+ Earl P.N. Duque. Facing the Knowledge Extraction and Visualization Challenges of the NASA CFD 2030 Vision. In _AIAA Aviation 2015_, Dallas, TX, USA, June 2015.
+ E. Wes Bethel. Brook No Delay: In Situ Visualization and Analytics on HPC Platforms. In _BIS 2015_, Inria@SiliconValley, Panel on Big Data Science: Data Analytics Meets High Performance Computing, Berkeley, CA, USA, May 2015.
+ Gunther H. Weber. Computing and visualizing time-varying merge trees for high-dimensional data. In _TopoInVis 2015_, Annweiler, Germany, May 2015. Best paper award.
+ Earl P.N. Duque. Accelerating the Post-Processing of Large Scale Unsteady CFD Applications via In Situ Data Reduction and Extracts. _Department of Aeronautical Engineering Graduate Seminar_, Embry-Riddle University, Daytona Beach, FL, USA, April 2015.
+ Dmitriy Morozov. Computing topology in parallel. In _Topological and Geometric Data Analysis seminar_, Ohio State University, Columbus, OH, April 2015.
+ E. Wes Bethel. Scientific Visualization. _Invited lecture_, MSIM 742, Synthetic Environments and Advanced Visualization, Old Dominion University, Norfolk, VA, USA, February 2015.
## 2014
+ Dmitriy Morozov. Distributed computation of persistent homology using the blowup complex. In _Computational Topology and Geometry workshop at the Foundations of Computational Mathematics (FOCM) conference_, Montevideo, Uruguay, December 2014.
+ Dmitriy Morozov. Wrinkles on everest: Persistence and stability in an omniscalar world. In _Department of Computer Science colloquium_, Tulane University, New Orleans, LA, November 2014.
<!-- extra line breaks to prevent footer from obscuring text -->
<br><br><br>
# Software Design
SENSEI provides simulations with a generic data interface that they use to provide access to their state. SENSEI then passes this data to zero or more analysis and visualization tasks each time the simulation provides more data.
## Generic Data Interface
There are two main challenges to using in situ analysis for advanced modeling and simulation workflows. First, on the simulation side, is the complexity of instrumenting simulation codes to use any in situ infrastructure. Presently, one has to instrument their simulation codes separately for each of the infrastructures. Each infrastructure has their own idiosyncrasies that the application developer has to endure, including mapping simulation data structures to the target infrastructure. Second, on the analysis side, analysis developers face the challenge of having to decide on the infrastructure in which to implement their analysis. It is not feasible to write analysis code once and use it in various infrastructure without modifications.
The SENSEI generic data interface addresses both these key challenges. First, it provides application developers with a generic data interface that they then tailor for a particular use. Second, it provides analysis developers with a data model that they may use to write analysis routines. Both of these components are independent of the in situ infrastructure being used and hence provide both the simulation and the analysis routine isolation from which in situ infrastructure is being used. For example, if the application is instrumented with the SENSEI interface, application end-users can easily choose between ParaView/Catalyst and VisIt/Libsim for generating visualizations in situ. Furthermore, since ParaView/Catalyst and VisIt/Libsim both are treated as analysis routines under SENSEI, these visualizations can be run in situ, or in transit using ADIOS or GLEAN transparently.
This write once, use anywhere goal is only achievable when we have a mutually agreed platform for communicating the data between the simulation and analysis components – the data model. For the SENSEI interface, we selected the VTK data model. The VTK data model is widely used in the scientific and engineering data analysis and visualization community, leveraged by visualization tools like ParaView and VisIt and hence already familiar to a broader community.
To minimize effort and memory overhead when mapping memory layouts for data arrays from applications to VTK, we enhanced the VTK data model to support arbitrary layouts for multi-component arrays. VTK now natively supports the commonly encountered structure-of- arrays and array-of-structures layouts. This allows for mapping data arrays from application codes to the VTK data model without additional memory copying (zero-copy).
Besides the data model, the other components that comprise the SENSEI interface are simple and quite light weight. The figure shows the main components of the SENSEI interface. The data adaptor provides a mapping between simulation data structures and the VTK data model. The analysis adaptor passes the data described in form of VTK data objects to any analysis code, doing any necessary transformations. The in situ bridge is a simple mechanism to put together the analysis workflow i.e. initialize the data adaptor and execute selected analysis routines.
To instrument an application with SENSEI, one provides a concrete implementation for the data adaptor API. The data adaptor API provides the analysis code with access to mesh and attributes arrays as needed. By providing an API that encourages lazy mapping to VTK data model for the mesh and attribute arrays, the data adaptor avoids any work to map simulation data to VTK data when not needed. Thus when no analysis is enabled, the SENSEI instrumentation overhead is almost nonexistent.
To add an analysis routine to SENSEI, one provides a concrete implementation for the analysis adaptor API. The analysis adaptor is provided an instance of the data adaptor that it may use to gain access to the simulation data through VTK data model.
Finally, the in situ bridge is simply an API and the corresponding implementation that the application developer implements to pass data and control to SENSEI during the application execution. A typical bridge implementation will initialize the data adaptor and one or more analysis adaptors during the initialization phase of the simulation; then for each time step pass the current simulation data arrays and any other metadata to the data adaptor and call execute on the analysis adaptors.
The analysis adaptor is also the mechanism for the SENSEI interface to connect with the different in situ infrastructures. For example, an analysis adaptor may use ADIOS to save the data out to an ADIOS BP file, or it may serve as a ParaView/Catalyst-based adaptor that starts up ParaView/Catalyst to process the data using ParaView/Catalyst data processing pipelines, including rendering.
The SENSEI generic data interface creates several possibilities for in situ, in transit, in flight and hybrid analysis. In enables a developer to instrument a simulation code once, then have access to multiple in situ infrastructures. Allowing additional in situ infrastructures to be coupled via the SENSEI generic data interface provides a number of analysis techniques to map to future high-performance computing architectures.
The current limitations of the SENSEI interface are an incomplete data model and an immature analysis adaptor specification. The SENSEI interface will truly be simpler when more complex simulation data structures easily map to the SENSEI data model through the data adaptor. Although this study examined several analysis and visualization use cases, this is just the tip of an iceberg of analysis techniques, and the adaptor infrastructure must grow to accommodate the requirements of the others.
Download the SENSEI generic data interface [here](https://gitlab.kitware.com/sensei/sensei).
## Analysis and Visualization
**ParaView Catalyst** (aka Catalyst) is an in situ analysis and visualization library that enables using ParaView’s visualization capabilities in in situ workflows. Applications can use Catalyst to execute complex analysis pipelines in step with the simulation, as well as connecting with the ParaView GUI for live, interactive visualization. To minimize memory footprint, Catalyst libraries are available in various flavors, called editions, that only enable components of ParaView used in the analysis pipelines.
[http://www.paraview.org/in-situ/](http://www.paraview.org/in-situ/)
**Libsim** is a library that makes available the full complement of features from VisIt so they may be used in situ. Libsim enables VisIt to connect interactively to running simulations for live exploration. Libsim can also be used directly to set up visualizations or it can use VisIt session files, which are XML files saved from the VisIt GUI, which can specify more complex visualizations. Once visualizations are set up, Libsim can save images for movie-making or it can save reduced-size data extracts for post hoc analysis.
**GLEAN** is a flexible and extensible framework that takes application, analysis, and system characteristics into account to facilitate simulation-time data analysis and I/O acceleration. The GLEAN infrastructure hides significant details from the end user, while at the same time providing a flexible interface to the fastest path for their data and analysis needs and, in the end, scientific insight. It provides an infrastructure for accelerating I/O, interfacing to running simulations for in transit analysis, and/or an interface for in situ analysis with zero or minimal modifications to the existing application code base.
**ADIOS** is an adaptive I/O service that is designed to allow applications to easily change between different I/O service providers. Only a tweak to the input parameters is needed to swap methods. This design allows for rapid conversion of post hoc analysis pipelines to in situ, in transit, or hybrid solutions by using one of the memory-to-memory “staging” methods, such as FlexPath or DataSpaces. The Flex-Path transport used in this effort can support same-node, multi-node, or even multi-machine deployment configurations. Unlike Catalyst and Libsim, however, ADIOS does not include any of the analytics functionality itself; it marshals the memory and metadata to make such code self-describing and adaptable to new situations. As such, it can partner effectively with Catalyst, Libsim, and other analytics infrastructures to provide whatever tools the scientist currently needs.
<!-- extra line breaks to prevent footer from obscuring text -->
<br><br><br>
This diff is collapsed.
learn-more: /learn-more/
develop: /develop/
news: /news/
menu:
learn-more: Learn more
develop: Develop
news: News
index:
learn-more: Learn more
develop: Develop
news: News
page:
contents: Contents
back_to_top: Back to Top
improve: Improve this doc
prev: Prev
next: Next
last_updated: "Last updated: %s"
sidebar:
learn-more:
project: Project
about: About
overview: Overview
presentations: Presentations
software: Software
develop:
using: Using
overview: Overview
dataset-schema: Dataset schema
testing: Testing
contributing: Contributing
learn-more:
project:
overview: /sensei/learn-more/index.html
about: /sensei/learn-more/about.html
presentations: /sensei/learn-more/presentations.html
software: /sensei/learn-more/software.html
develop:
using:
overview: /sensei/develop/index.html
dataset-schema: /sensei/develop/adios-schema.html
contributing:
overview: /sensei/develop/contribute.html
testing: /sensei/develop/testing.html
{
"name": "sensei",
"version": "0.0.0-semantically-release",
"description": "in situ framework",
"keywords": [
"3d",
"visualization",
"sensei",
"vtk",
"graphics"
],
"repository": {
"type": "git",
"url": "git+https://gitlab.kitware.com/sensei/sensei.git"
},
"license": "BSD-3-Clause",
"bugs": {
"url": "https://gitlab.kitware.com/sensei/sensei/issues"
},
"homepage": "https://gitlab.kitware.com/sensei/sensei",
"devDependencies": {
"kw-doc": "1.1.3",
"kw-web-suite": "6.0.2"
},
"scripts": {
"doc": "kw-doc -c ./doc/config.js",
"doc:www": "kw-doc -c ./doc/config.js -s",
"doc:publish": "kw-doc -c ./doc/config.js -mp"
}
}
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment