Index  | Recent Threads  | Unanswered Threads  | Who's Active  | Guidelines  | Search
 

Quick Go »
No member browsing this thread
Thread Status: Active
Total posts in this thread: 5
[ Jump to Last Post ]
Post new Thread
Author
Previous Thread This topic has been viewed 1039 times and has 4 replies Next Thread
Former Member
Cruncher
Joined: May 22, 2018
Post Count: 0
Status: Offline
Reply to this Post  Reply with Quote 
anyone try this on a mainframe

Just out of curiosity, has anyone set up a big mainframe running VM with several hundred / thousand linux systems and run these wu's to see what kind of performance is derived?

I'm just curious how fast these might run on say a big z9 processor with numerous LPARS and VM systems with lots of linux guests.

Anyone have this data or know where to get it?

TIA
[Jul 26, 2008 2:02:07 PM]   Link   Report threatening or abusive post: please login first  Go to top 
Former Member
Cruncher
Joined: May 22, 2018
Post Count: 0
Status: Offline
Reply to this Post  Reply with Quote 
Re: anyone try this on a mainframe

Most of the science applications we use were originally used on supercomputers. I have no idea what information is available - do let us know what you find out.
[Jul 26, 2008 2:50:23 PM]   Link   Report threatening or abusive post: please login first  Go to top 
Dotsch
Advanced Cruncher
Joined: Feb 12, 2006
Post Count: 100
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: anyone try this on a mainframe

I have build a SETI application for Linux on the zSeries. But, I have to look for the excatly perofmance benchmarks I have done. I think it was in the range of a midrange PC. - The zSeries is optimzed for a heavy workload with a heavy throughtput and not specialy for fast numbercrunching.

So far I know has IBM announced, or launched, Cell CPU based boards for the zSeries for special tasks. If you rewritte the science application specialy for the Cell CPU (like it also would be used in the Playstation 3 (PS3) you will get a very good performance.
There are several BOINC project out there which got a really good perfomance from the Cell CPU...
[Aug 13, 2008 9:29:26 PM]   Link   Report threatening or abusive post: please login first  Go to top 
akamai99
Cruncher
Joined: Aug 11, 2005
Post Count: 4
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: anyone try this on a mainframe

I have lots of empty Cycles on some Mainframes, but I don't know how to build these. Is there a set of instructions or someone interested in building these for me... biggrin
[Aug 20, 2008 5:35:21 AM]   Link   Report threatening or abusive post: please login first  Go to top 
Dotsch
Advanced Cruncher
Joined: Feb 12, 2006
Post Count: 100
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: anyone try this on a mainframe

Porting BOINC is quite easy on Linux for zSeries. A BOINC client binary package and the SETI application is still available at my homepage (http://www.dotsch.de).
But the problem at WCG and the most other projects is, that they are closed source. There are only few BOINC projects, which are open source...

What excatly do you want to port ? - Eventualy I can assist you...
----------------------------------------
[Edit 2 times, last edit by Dotsch at Aug 20, 2008 3:41:05 PM]
[Aug 20, 2008 3:38:55 PM]   Link   Report threatening or abusive post: please login first  Go to top 
[ Jump to Last Post ]
Post new Thread