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: 86
Posts: 86   Pages: 9   [ Previous Page | 1 2 3 4 5 6 7 8 9 | Next Page ]
[ Jump to Last Post ]
Post new Thread
Author
Previous Thread This topic has been viewed 14098 times and has 85 replies Next Thread
macusen
Advanced Cruncher
Germany
Joined: Mar 31, 2009
Post Count: 59
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: The Road to 600 TeraFLOPS: Need More Powah to stay Smarter Than Watson.

I think we met the 600 TFL the last days biggrin

Good work, always keep growing! smile
----------------------------------------

[Nov 22, 2011 2:36:26 PM]   Link   Report threatening or abusive post: please login first  Go to top 
Hypernova
Master Cruncher
Audaces Fortuna Juvat ! Vaud - Switzerland
Joined: Dec 16, 2008
Post Count: 1908
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: The Road to 600 TeraFLOPS: Need More Powah to stay Smarter Than Watson.

If we reach 1 PetaFlops through a volountary computing grid, I think that it will be a great success and a major milestone. The next generation CPU's from Intel (more cores, less power), should help.
Next week I will upgrade one of my rigs, with a new Intel X3960 CPU smile . Let's see what it is capable of biggrin . Stay tuned.
----------------------------------------

[Nov 22, 2011 10:30:18 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: The Road to 600 TeraFLOPS: Need More Powah to stay Smarter Than Watson.

This morning's numbers suggest we're first time scaling the 400 CPU years in a day... could be as high as 410-415, if the servers hold together [knreed is sequestered through the early hours to make sure :O]. With that comes 640-650 TFL and the machine been humming at 21 day mean at 560 TFL.

Crunching On

--//--


[Nov 28, 2011 2:01:35 PM]   Link   Report threatening or abusive post: please login first  Go to top 
Pink Moose
Senior Cruncher
Joined: Nov 19, 2010
Post Count: 170
Status: Offline
Reply to this Post  Reply with Quote 
Re: The Road to 600 TeraFLOPS: Need More Powah to stay Smarter Than Watson.

I keep all three of my computers-hyperthreading and running 24/7 for the WCG. I really like the intel i7 CPU's. It's nice to be able to run 8 jobs at a time off of one computer that has an intel i7 CPU in it.

If we can all keep spreading the word around about our World Community Grid, in time, we'll "hit higher and higher marks" as more and more people understand the importance of contributing computer time to save lives world wide.
[Nov 28, 2011 9:39:54 PM]   Link   Report threatening or abusive post: please login first  Go to top 
Hypernova
Master Cruncher
Audaces Fortuna Juvat ! Vaud - Switzerland
Joined: Dec 16, 2008
Post Count: 1908
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: The Road to 600 TeraFLOPS: Need More Powah to stay Smarter Than Watson.

I know that this is an eternal subject. But still I must admit that it is very disturbing that between our measure of performance and the one posted for the WCG project in boincstats there is a factor of two. I know the way we and they measure but whatever the explanations a factor of two is so large that both values are probably wrong an difficult to be trusted.
Who is nearer to the real performance ? shock
----------------------------------------

[Dec 1, 2011 4:27:06 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: The Road to 600 TeraFLOPS: Need More Powah to stay Smarter Than Watson.

Dr. Anderson decided last year to change the policy of the performance measure and were to be halved and is now using a 200 divisor instead of 100 (Cobblestones). I've not attempted to understand and me and SNURK have stuck to the old way in charts and sigs. This is the relative way of measuring past and present performance, here.

--//--

edit: And our 60 day average just passed 550 TFL :P
----------------------------------------
[Edit 2 times, last edit by Former Member at Dec 1, 2011 4:45:13 PM]
[Dec 1, 2011 4:42:36 PM]   Link   Report threatening or abusive post: please login first  Go to top 
Hypernova
Master Cruncher
Audaces Fortuna Juvat ! Vaud - Switzerland
Joined: Dec 16, 2008
Post Count: 1908
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: The Road to 600 TeraFLOPS: Need More Powah to stay Smarter Than Watson.

Ok. So did I too in my signature to keep a relative coherence.
But I think it is a subject that has to be studied one day in depth.
----------------------------------------

[Dec 1, 2011 6:19:02 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: The Road to 600 TeraFLOPS: Need More Powah to stay Smarter Than Watson.

We did 667 TFL yesterday, bringing the "trimmed" 21 day mean to 590.4 TFL. The 60 days average stats are at 560 now. We're getting there, and maybe even before the new year. :D

Crunch On.

--//--
[Dec 13, 2011 8:12:54 AM]   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: The Road to 600 TeraFLOPS: Need More Powah to stay Smarter Than Watson.

Strictly speaking, we yesterday passed the 600TFL on 21 day average, but due the tech work performed delaying about 60 CPU years of validations to today it did not happen... then, because the 60 years gets added onto the 15th, today got marked as an outlier too... chucked out of the trimmed mean. Maybe Friday we can say so with an honest face, that we have "The Arrival", and assume building the expansion to 700TFL machine.

Crunch On.

--//--
----------------------------------------
[Edit 1 times, last edit by Former Member at Dec 15, 2011 5:19:34 PM]
[Dec 15, 2011 5:18:31 PM]   Link   Report threatening or abusive post: please login first  Go to top 
E. Frijters
Senior Cruncher
The Netherlands
Joined: Apr 26, 2007
Post Count: 228
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: The Road to 600 TeraFLOPS: Need More Powah to stay Smarter Than Watson.

Way to go Rob!

Crunch on!
----------------------------------------
Former grid.org slave


[Dec 17, 2011 9:09:15 PM]   Link   Report threatening or abusive post: please login first  Go to top 
Posts: 86   Pages: 9   [ Previous Page | 1 2 3 4 5 6 7 8 9 | Next Page ]
[ Jump to Last Post ]
Post new Thread