Skip to main content
UNIVERSITY OF ILLINOIS URBANA-CHAMPAIGN
Toggle navigation
Research
Dropdown menu toggle
Artificial Intelligence
Arts and Humanities
Astrophysics
Digital Agriculture
Earth and Environment
Engineering
Health Sciences
Project Highlights
Expertise
Dropdown menu toggle
Compute Resources
Data Analytics
Facilities
Innovative Systems
Integrated Cyberinfrastructure
Program Administration
Software and Applications
User Services
Visualization
People
Dropdown menu toggle
Leadership
Directorates
Staff Directory
News & Events
Dropdown menu toggle
News
Calendar
Press Room
Tours
About
Dropdown menu toggle
Careers
Fellowships & Internships
Industry Partner Program
Institutional Partnerships
Diversity
History
Giving
Contact
Search
Search
Toggle navigation
Calendar
National Center for Supercomputing Applications WordPress Master Calendar
Share on Facebook
Tweet
Email
add to calendar
contact
add an event
View Full Calendar
NCSA staff who would like to submit an item for the calendar can email
newsdesk@ncsa.illinois.edu
.
Efficient I/O: HDF5 and DLIO Workshop
Event Type
Conference/Workshop
Sponsor
NCSA
Date
Mar 12, 2024 1:00 - 3:00 pm
Registration
Registration form
Contact
Bruno Abreu
E-Mail
babreu@illinois.edu
Views
19
Originating Calendar
NCSA Research Consulting Training Events
This workshop will include two talks about optimizing, profiling, and benchmarking input/output operations using
HDF5
and
DLIO
.
Optimizing Your I/O Workload: Techniques for Effective HDF5 Usage
- Scot Breitenfeld, The HDF Group
HDF5 is a widely used data model, file format, and I/O library, particularly in HPC applications for managing and storing large amounts of simulation data. This talk will focus on HDF5 usage on NCSA's Delta system and provide a brief overview of HDF5 (serial and parallel) with an emphasis on HDF5 HPC tuning techniques such as collective metadata I/O, data aggregation, asynchronous I/O, and other HDF5 tuning parameters and features. During the presentation, we will discuss various storage options available on Delta for HDF5 files and multiple post-simulation storage options, including cloud storage and other tools.
Deep Learning I/O: Benchmark and Profiling
- Hariharan Devarajan, Lawrence Livermore National Laboratory
In this talk, we will be taking a dive into the unique features of I/O in deep learning application. We will present our ongoing efforts to understand the I/O characteristics of DL workloads using our DLIO profiler. Finally, we will present the DLIO Benchmark which is built to accurately represent the I/O characteristics in deep learning workloads.
link for robots only
Back to top