Difference between revisions of "XC-40 Architecture (PDC February 2015)"
(Created page with "{{training event info |description=Using Cray XC-40 Machines |fields=Scientific computing |event type=Course |location=Stockholm |start date=2015-02-03 |end date=2015-02-06 }} ==...") |
(→Trainer) |
||
| (4 intermediate revisions by 2 users not shown) | |||
| Line 8: | Line 8: | ||
}} | }} | ||
== Seminar overview == | == Seminar overview == | ||
| − | + | This workshop (which runs over four days, starting and ending at lunchtime) is for people who use, or are interested in using, Cray XC supercomputers - such as Milner and Beskow at PDC. (Note that during 2014 Cray changed the name of the XC30 system to XC40 - so at PDC, Milner was delivered as an XC30 machine while Beskow is an XC40 system.) The workshop will be run by staff from Cray. You will learn how to efficiently port, optimize and execute applications on the Cray XC systems at PDC. The workshop is a mixture of lectures and practical sessions. Example exercises will be provided but attendees are strongly encouraged to bring their own applications to the workshop, as you will have the opportunity to tune your own applications during the hands-on labs with help from Cray staff. | |
| + | |||
| + | The main focus of the course is on porting and optimizing your applications on Cray systems. More specifically, the course will cover the Cray XC30 and XC40 architecture and programming environment, and give a basic overview of scientific libraries. | ||
| + | |||
| + | In addition there will be a focus on identifying performance bottlenecks in scientific applications and potential causes using the Cray Performance tools CrayPat, Cray Reveal and Cray Apprentice2 for profiling and visualization. | ||
| + | |||
| + | Finally there will be opportunity to learn various optimization techniques related to single CPUs and input/output, along with giving an overview of MPI3 and PGAS, as well as Coarray Fortran and Unified Parallel C. | ||
| + | |||
| + | The course also includes practical work, ideally on your own code which you should bring with you. The focus of the practical work will be on porting applications to the XC-30/XC-40 platform and there will be a short introduction to debugging tools to help resolve any problems with the initial port. | ||
| + | |||
| + | More information can be found on the [https://www.pdc.kth.se/events/event-repository/using-cray-xc40-and-xc30-systems-february-2015 PDC web page] | ||
== Trainer == | == Trainer == | ||
| − | The course | + | The course will be run by Cray Staff |
Latest revision as of 17:07, 4 December 2014
| Name | XC-40 Architecture (PDC February 2015) |
|---|---|
| Description | Using Cray XC-40 Machines |
| Type of event | Course |
| Location | Stockholm |
| Start date | 2015-02-03 |
| End date | 2015-02-06 |
Seminar overview
This workshop (which runs over four days, starting and ending at lunchtime) is for people who use, or are interested in using, Cray XC supercomputers - such as Milner and Beskow at PDC. (Note that during 2014 Cray changed the name of the XC30 system to XC40 - so at PDC, Milner was delivered as an XC30 machine while Beskow is an XC40 system.) The workshop will be run by staff from Cray. You will learn how to efficiently port, optimize and execute applications on the Cray XC systems at PDC. The workshop is a mixture of lectures and practical sessions. Example exercises will be provided but attendees are strongly encouraged to bring their own applications to the workshop, as you will have the opportunity to tune your own applications during the hands-on labs with help from Cray staff.
The main focus of the course is on porting and optimizing your applications on Cray systems. More specifically, the course will cover the Cray XC30 and XC40 architecture and programming environment, and give a basic overview of scientific libraries.
In addition there will be a focus on identifying performance bottlenecks in scientific applications and potential causes using the Cray Performance tools CrayPat, Cray Reveal and Cray Apprentice2 for profiling and visualization.
Finally there will be opportunity to learn various optimization techniques related to single CPUs and input/output, along with giving an overview of MPI3 and PGAS, as well as Coarray Fortran and Unified Parallel C.
The course also includes practical work, ideally on your own code which you should bring with you. The focus of the practical work will be on porting applications to the XC-30/XC-40 platform and there will be a short introduction to debugging tools to help resolve any problems with the initial port.
More information can be found on the PDC web page
Trainer
The course will be run by Cray Staff
