Parallel Computing on Stampede

Event Date: 

Thursday, January 10, 2013 - 8:30am to Friday, January 11, 2013 - 5:00pm

Description: 

ITS-Research Services has reserved a room for the viewing of the Parallel Computing on Stampede training class presented by the Texas Advanced Computing Center.

Stampede is the newest national supercomputing resource run by the Texas Advanced Computing Center and features the new Xeon Phi coprocessor architecture. University of Iowa users can access this system, and other resources, through the Xsede program.

Please register with TACC as discussed below and fill out the short form on this page if you plan to attend the webcast in 2837 UC (if viewing from another location you do not need to register on this page).

Class Information:

January 10-11, 2013 (Thursday and Friday)

8:30 a.m. to 5 p.m. CT, both days

You are welcome to attend this training class in-person or via webcast.



Registration closes January 3.

 

Please bring your laptop and ensure you have an ssh client installed.


The MIC instruction will be the second half of the second day of the class, Friday, January 11.


The labs will be available for remote users.  However, we will not be able to assist remote users with problems during the lab.  Remote users are invited to submit their questions via email.

 

Agenda  (Agenda is subject to change)



Day 1 (Thursday, Jan 10, 2013)

 



Morning (Basics)



08:30-08:45 TACC Overview


08:45-09:45 Introduction to Parallel Computing


09:45-10:00 Break


10:00-11:00 User Environment


11:00-11:30 Lab 1 (User Environment)


11:30-12:30 Lunch
 

 


Afternoon (Programming)



12:30-13:50 OpenMP Programming


13:50-14:30 Lab 2 (OpenMP)


14:30-14:45 Break


14:45-16:10 MPI Programming


16:10-17:00 Lab 3 (MPI)

 
Day 2 (Friday, January 11, 2013)

 


Morning (Advanced)

 

08:30-09:30 TACC Systems Review


09:30-10:30 Optimization & Scalability


10:30-10:45 Break


10:45-11:30 Hybrid Computing


11:30-12:00 Lab 4 (Hybrid)


12:00-12:45 Lunch
 


Afternoon (Coprocessing with MIC)

 



12:45-13:30 MIC (Xeon Phi Coprocessor)


13:30-14:30 Native Computing (& Optimization)


14:30-14:45 Break


14:45-15:15 Symmetric Computing


15:15-16:15 Offloading


16:15-17:00 Lab 5 (MIC)

 

Event Location: 

2837 UCC
US