Weekly Update: March 27, 2019
IN THIS MESSAGE
- Center Announcements
- Notable System Changes in 2019
- HPSS Now Accessible Via Globus
- 2019 OLCF User Meeting (May 21-23)
- Meetings & Workshops
- March OLCF User Conference Call (Mar 27)
- VI-HPS Tuning Workshop (Apr 9-12)
- Introduction to NVIDIA Profilers on Summit (Apr 11)
- Quantum Computing User Forum (Apr 25)
- Introduction to Summit Workshop (May 20)
- OLCF GPU Hackathon Series (multiple dates)
- Upcoming Scheduled Outages
- None through April 5.
NOTABLE SYSTEM CHANGES IN 2019
On February 25, 2019, we notified users of several notable system changes coming in 2019. The last day to submit jobs to Titan and Eos will be June 30th. Atlas will be made read-only on August 1 and all data will be permanently deleted on August 15. We strongly encourage users to begin moving data out of that filesystem and onto the newer Alpine filesystem or to some alternate locate as soon as possible. If you have questions about data movement, please contact firstname.lastname@example.org. For more information, please see the 2019 OLCF System Decommissions and Notable Changes page on the OLCF website.
HPSS NOW ACCESSIBLE VIA GLOBUS
We are happy to announce that users can now use Globus to transfer files to our High Performance Storage System (HPSS). HPSS is available via the “OLCF HPSS” endpoint. While our tests have shown that it is sometimes more efficient to transfer via Globus than hsi/htar, there are several restrictions of which you should be aware: there is a limit of 8 active transfers across all users and each user has a limit of 3 active transfers. You should transfer large amounts of data over a few connections rather than small amounts of data over many connections. For directories with large numbers of small files, you should first create a .tar file of that directory and then transfer the .tar file. More information about this capability is available on the 2019 OLCF System Decommissions and Notable Changes page.
2019 OLCF USER MEETING
The Oak Ridge Leadership Computing Facility (OLCF) is pleased to announce the 2019 OLCF User Meeting will be held May 21-23 at Oak Ridge National Lab in Oak Ridge, TN. The purpose of the annual user meeting is to share selected computational science and engineering achievements emerging from OLCF’s user programs, to enable direct interactions among users, advance OLCF’s relationships with our user community, and highlight computational requirements for the future. Highlights of the meeting include paired talks from OLCF users and application developers, plenary talks, updates from center staff, a poster session, OLCF User Group Executive Board elections, and tours of OLCF. Additionally, OLCF will offer training on the Monday prior to the meeting for those interested. For more information or to register, please visit the event page.
MEETINGS & WORKSHOPS
MARCH OLCF USER CONFERENCE CALL (MAR 27)
The March OLCF User Conference Call will be held at noon (Eastern
Time) on Wednesday, March 27. The call will include a discussion
of CUDA 10 features by Stephen Abbott of NVIDIA. Details of the
call, including information on how to join, can be found at
VI-HPS TUNING WORKSHOP (APRIL 9-12)
The Virtual Institute – High Productivity Supercomputing (VI-HPS) is holding a tuning workshop at the Innovative Computing Laboratory (ICL) at the University of Tennessee (UTK) from April 9-12, 2019. The goal of the workshop is to give an overview of the VI-HPS tools suite (Arm tools, PAPI, Paraver, Score-P, etc.) and help participants to analyze and tune their own MPI, OpenMP, and hybrid MPI+OpenMP parallel applications. TACC’s Stampede2 will be the primary platform for the workshop, and course accounts will be provided during the workshop to participants. For more information, or to register, please visit the event page.
INTRODUCTION TO NVIDIA PROFILERS ON SUMMIT (APR 11)
The Oak Ridge Leadership Computing Facility (OLCF) will host an
Introduction to NVIDIA Profilers tutorial on April 11, 2019. This
event is intended to teach OLCF users a practical approach to
using the NVIDIA profilers on Summit. Participants will be guided
through example programs to demonstrate the recommended workflow
for using the NVIDIA profilers on Summit. The topics covered will
include an introduction to the text-based and visual (GUI)
profilers, how to handle single- and multi-process profiling, and
basic customization/annotation of visual profiling with the NVTX
API. For more information, or to register, please visit the event
page at: https://www.olcf.ornl.gov/calendar/introduction-to-nvidia-profilers-on-summit/
QUANTUM COMPUTING USER FORUM (APR 25)
ORNL is hosting a Quantum Computing User Forum on April 25. This meeting will bring together users to discuss common practices in the development of applications and software for quantum computing systems. The one-day meeting will include invited and contributed presentations from leading researchers in the field of quantum computer science, and it will provide a forum in which to discuss the latest scientific applications of quantum computers. An optional training event will be held one day prior to the Forum. For more information, visit the event page.
INTRODUCTION TO SUMMIT WORKSHOP (MAY 20)
The OLCF will host an Introduction to Summit workshop on Monday, May 20, 2019 from 8:00 AM – 5:00 PM. The event will consist of presentations in the morning followed by a hands-on session in the afternoon. The presentations are meant to deliver relevant information about the system (hardware, job launcher, programming environment, etc.), while the hands-on session is intended to demonstrate specific aspects of the Summit system through a set of provided example programs. For more information, or to register, please visit the event page.
OLCF GPU HACKATHON SERIES (OPEN CALLS)
OLCF is once again supporting several GPU hackathons. GPU hackathons are 5-day coding events in which teams of developers prepare their own applications(s) to run on GPUs or focus on optimizing their application(s) that currently run on GPUs. Teams should consist of three or more developers who are intimately familiar with (some part of) their application, and they will work alongside two mentors with GPU programming expertise. If you want/need to get your code running (or optimized) on a GPU-accelerated system, these hackathons offer a unique opportunity to set aside 5 days, surround yourself with experts in the field, and push toward your goals. If you are interested in more information, or would like to submit a proposal to one of the upcoming events, please visit the OLCF GPU Hackathons event page.
Upcoming GPU Hackathons and application deadlines:
MIT: June 3-7 (deadline to apply: April 5)
Princeton: June 24-28 (deadline to apply: April 26)
UPCOMING SCHEDULED OUTAGES
None through April 5.