ITS Cyberinfrastructure News
Hawai‘i depends exclusively on local water. The ‘Ike Wai project will provide data and models that address the grand challenge of water sustainability. A diverse workforce of data scientists and water researchers will work in concert with the community, government and business to inform decision makers with high-quality data and predictive capacity. The project incorporates indigenous and local communities, and its robust, inclusive and diverse human capital pipeline of undergraduates, graduate students, postdocs and junior faculty will address water challenges at the academic and policy level. For more information, go to the Hawaiʻi EPSCoR website.
On April 13th the Texas Advanced Computing Center’s (TACC) Rion Dooley, John Fonner and Joe Stubbs conducted a workshop for UH researchers covering the use of the Agave platform (http://agaveapi.co) for supporting research data management and computation. The full day workshop walked researchers through using the Agave command line tools inside a Jupyter notebook as they registered the UH ITS HPC compute system and scratch file system in Agave and created some test applications that move data, run computation and archive results in an automated fashion.
UH and TACC are partners on a $4 million dollar NSF grant to bring this “Science As A Service” platform to UH and other institutions to support and accelerate research. Agave allows researchers to leverage existing storage and computation systems both locally, nationally and internationally as part of their project infrastructure. Agave does this by making it easier to share data and software with other researchers using the technologies we use everyday such as the web and as more tools arrive for the platform they will allow researchers to rely less on the command line and terminal.
Already there have been several online science gateways built using the Agave platform including: the Arabidopsis information portal , Araport which offers gene and protein reports with orthology, expression, interactions and the latest annotation, analysis tools, community apps, the iPlant Collaborative, now CyVerse, which provides life scientists with powerful computational infrastructure to handle huge datasets and complex analyses that enable data-driven discovery, BioExtract an open, web-based system designed to aid researchers in the analysis of genomic data by providing a platform for the creation of bioinformatic workflows, and DesignSafe is the cyberinfrastructure component of the NSF-supported Natural Hazards Engineering Research Infrastructure (NHERI).
Advancing Research Computing on Campuses (ARCC): Best Practices Workshop
March 22-24, 2016
This workshop is presented in partnership with Advanced Cyberinfrastructure Research and Education Facilitator (ACI-REF) institutions, which the University of Hawaii is a member, and the National Center for Supercomputing Applications (NCSA). Professionals involved in operating and supporting campus shared research computing infrastructure are invited to join in this workshop event and share experiences and expertise. Proposed topics include:
- Best practices for running advanced computing resources in a higher education environment
- The condo model
- Business models
- Collaboration with researchers
- Return on investment
- Interactions with national data centers and infrastructures
The workshop will be held March 23rd-24th in Champaign IL. Tutorial sessions will b offered on March 22nd. For more information and register please go to the main site – http://www.ncsa.illinois.edu/Conferences/ARCC/index.html
Event Details: NVIDIA, Institute for Astronomy and Cyberinfrastructure are pleased to be organizing a 2-day High Performance Computing and Programming event.
Event Location: University of Hawaii, Information Technology Center, Room ITC-105
Why you should attend: NVIDIA GPUs are the world’s fastest and most efficient accelerators delivering world record scientific application performance. NVIDIA’s CUDA Technology is the most pervasive parallel computing model, used by over 250 scientific applications and over 150,000 developers worldwide. This Programming Workshop will focus on introducing scientific computing programming utilizing NVIDIA GPUs to accelerate applications across a diverse set of domains.
Presented by NVIDIA instructor Dr. Jonathan Bentz, the workshop will introduce programming techniques using CUDA and OpenACC paradigms as well as optimization, profiling and debugging methods for GPU programming. An introduction to Deep Learning using GPUs will also be covered.
Who it’s for: Graduate Students, Postdocs, Researchers, and Professors
Agenda: February 4th Day 1: 9AM to 4:30PM
Introduction to GPU programming
• High Level Overview of GPU architecture
• OpenACC: An introduction on compiler directives to specify loops and regions of code in standard C, C++ and Fortran to be offloaded from a host CPU to an attached accelerator
• Hands-On examples to focus on data locality
• GPU-Accelerated Libraries: discussion including AmgX, cuSolver, cuBLAS and cuDNN
• Basics of GPU Programming; An introduction to the CUDA C/C++ Language
• 4 Hands-On examples will Illustrate simple kernel launches and using threads
Agenda: February 5th Day 2: 9AM to 4:30PM
Performance and Optimization
• Overview of Global and Shared memory usage
• Hands-On examples will illustrate a 1D Stencil and Matrix Transpose
• Using NVIDIA Profiler to identify performance bottlenecks
• Advanced Optimizations using Streams and Concurrency to overlap communication and computation
• Hands-On examples will use CUBLAS with Matrix Multiply
• Conclude with a Deep Learning Overview
Intro to Deep Learning / Machine Learning
• Overview of Global and Shared memory usage (Caffe, Torch, Theano)
• Deep Learning with GPUs and NVIDIA DIGITS
• Live demo using NVIDIA DIGITS
Caffe Lab Examples (time permitting)
Deep learning is a rapidly growing segment born from the fields of artificial intelligence and machine learning. It is increasingly used to deliver near-human level accuracy for image classification, voice recognition, natural language processing, sentiment analysis, recommendation engines, and more. Applications areas include facial recognition, scene detection, advanced medical and pharmaceutical research, and autonomous, self-driving vehicles. This overview will focus on introducing attendees to the use of GPU accelerated deep learning frameworks utilizing NVIDIA GPUs for ideal performance and scalability.
Coffee and Lunch will be provided
The following is taken from an article written by the Texas Advanced Computing Center. The original article can be found here( https://www.tacc.utexas.edu/-/accelerating-science-as-a-service-for-all-scientific-domains)
In today’s data-driven research environment, easy and reliable access to compute, storage, and data resources is as much a necessity as the algorithms used to make the actual discoveries. For U.S. researchers to stay competitive in the global research community and to reduce the time from theory to discovery, they need ready access to the digital resources that power their labs. In recent years, many researchers have turned to science gateways for such access.
On August 1, work officially started on the National Science Foundation (NSF)-funded Agave Platform, an open ‘Science-as-a-Service’ (ScaaS) cloud platform for reproducible science. The four-year grant totaling just under $4 million will leverage the success of what originated in 2011 as a set of hosted, web-based application programming interfaces (APIs) for plant biologists under the NSF iPlant project. Since that time, Agave has evolved into a viable software platform for researchers and developers in all domains of science.
“With this award, we will answer the question of how to easily and affordably put together a web-based infrastructure for a sustainable, on-demand and interactive gateway,” said Rion Dooley, a research associate at TACC and principal investigator (PI) of a new grant to explore this area.
Dooley says that everyone with an email account should be familiar with the Science-as-a-Service model, but that’s not quite the right fit for the research community. “The needs of researchers don’t fit into a tidy little box,” Dooley said. “We believe the Agave Platform will solve many of these issues by quickly closing the capability gap between academic and commercial infrastructure. This is our way to solve the problem for the computational science sector. We want this to be a web platform that helps the entire open science community realize their research goals and allows their ideas to be heard.”
Rajiv Ramnath, a program director in the Division of Advanced Cyberinfrastructure and the cognizant program officer for the Agave Platform said, “The project will enable the academic and research community to avail itself of what should become standard technology.”
He added, “The project team is made up of skilled developers, communicators and educators, and I look forward to successful outcomes from the project.”
While the Agave APIs were originally developed to power cyberinfrastructure for the plant biology community through iPlant, they have gained steady traction with researchers in other communities searching for simple, affordable ways to enable, accelerate and reproduce their science. Growing steadily since its initial release, the community driven design of the APIs allow Agave to deliver tangible value to nearly 20,000 researchers across five continents.
Building upon the success of the existing Agave Developer APIs, this new project includes three efforts, which will synergistically evolve the current technology into a sustainable ScaaS platform for the national research community:
- The Agave Platform APIs — The existing Agave developer APIs will expand to offer additional science APIs and management interfaces leading to a cohesive, self-provisioning platform that will enable reproducible ScaaS for the developer community and scale the platform to support broad usage from the national research community.
- The Application Exchange (AX) — The project team will partner with commercial and academic institutions to create a community-driven Application Exchange (AX) based on Docker container technology to facilitate application transparency, portability, attribution and reproducibility. The AX presents a highly interoperable, industry approved solution for creating and sharing reproducible science. By leveraging a matching industrial investment, this project will accelerate existing community efforts to create a culture of accountability and openness. This will enable a higher quality and frequency of peer reviewed research and more verifiable science.
- Agave ToGo — The project will consolidate existing open source contributions from projects already with the Agave ecosystem into Agave ToGo, a collection of reference science gateways in multiple languages and web frameworks. When combined with the Agave Platform and the AX, Agave ToGo will enable novice users to create scalable, reproducible, digital labs that span their office, commercial cloud and national data centers in a matter of minutes.
“Agave ToGo will make it easy to deploy science on the web,” Dooley said. “Users will be empowered to focus on domain science rather than computer science. Developers will stimulate innovation and increase the opportunity for discovery.” The cumulative effect of these efforts — combined with the ongoing education and outreach activities in classrooms, coding competitions, workshops and hackathons — will equip a new generation of students with a skillset highly coveted by industrial and academic employers alike.
Aside from Docker Inc., the creators of the container technology behind the AX, the industry partners include Appsoma, a popular science gateway used in the commercial and academic community, and The Center for Open Science, committed to providing ongoing support for the AX portal. In addition, the project will receive guidance from community development and technology transfer experts at Chapman University and The University of Texas at Austin.
“We’re really excited about these partnerships and to build this platform that the community can take and run with,” Dooley said.
In addition to Dooley, the project team includes John Fonner (Co-PI) from the Texas Advanced Computing Center; Gwen Jacobs (Co-PI) from the University of Hawai’i; and Jim Lupo (Co-PI) and Kathy Traxler from Louisiana State University. These team members bring decades of experience in API and science gateway development, community building, workforce development and education. Importantly, they provide a track record of working at the forefront of advanced computing and providing innovative, open source solutions to the academic research community.
The full original article can be found here (https://www.tacc.utexas.edu/-/accelerating-science-as-a-service-for-all-scientific-domains)