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: 9
[ Jump to Last Post ]
Post new Thread
Author
Previous Thread This topic has been viewed 2115 times and has 8 replies Next Thread
Former Member
Cruncher
Joined: May 22, 2018
Post Count: 0
Status: Offline
Reply to this Post  Reply with Quote 
Average process power? Gflops??

I was wondering is there a info how fast this grid is?? Im curious. It doesn't need to be exact and 100% correct. Who could know this thing?
[Jul 28, 2005 8:29:30 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: Average process power? Gflops??

one of the members of "my online team" posted an estimate the other day of 58 teraflops... here, about 7 posts down biggrin
[Jul 28, 2005 8:55:09 PM]   Link   Report threatening or abusive post: please login first  Go to top 
David Autumns
Ace Cruncher
UK
Joined: Nov 16, 2004
Post Count: 11062
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: Average process power? Gflops??

Hi musicmann,

Yesterday the WCG posted 12,850,066 points thats 535419 points per hour.

My PC scores 29.4 points per hour and benchmarks at 3.447 GFlops

So every hour there are 18211 PC's like mine crunching through Work Units for the WCG which would make 62775GFlops of raw number crunching power or 62.775 TeraFlops

This is the raw power of the WCG at the moment and would put is in 3rd place on the list of the top 500 supercomputer as of June this year.

There is redundancy however built into the system so that Work Units can be verified against each other due to the distributed nature of the Grid and the medical science that is being carried out here. I believe that each Work Unit is sent out about 5 times.

So the actual like for like comparison with a true Supercomputer would be around 12.5 TeraFlops give or take a GFlop or 2 which would still put the WCG a P14 in the list of the fastest supercomputers just 8 1/2 Months after the projects launch.

Which I think is just cool

take care and happy crunching

Regards
Dave
----------------------------------------

[Jul 28, 2005 8:58:16 PM]   Link   Report threatening or abusive post: please login first  Go to top 
David Autumns
Ace Cruncher
UK
Joined: Nov 16, 2004
Post Count: 11062
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: Average process power? Gflops??

Cheers Mousie
----------------------------------------

[Jul 28, 2005 8:58:56 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: Average process power? Gflops??

very nice you get the estimates, david smile
[Jul 28, 2005 9:02:25 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: Average process power? Gflops??

curious...is the redundancy set to five here? some folks over at grid.org mentioned it was only three. just wondering. smile
[Jul 28, 2005 9:04:38 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: Average process power? Gflops??

very good news.
Thanks for sharing that David.
[Jul 28, 2005 9:09:40 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: Average process power? Gflops??

Hello mousie,
curious...is the redundancy set to five here? some folks over at grid.org mentioned it was only three. just wondering. smile


This is something that is up in the air. A long time ago I posted a reply in which I used the figure 3. But later when I started searching, that was the only post I could find. I cannot remember what gave me that idea, so I have to assume that I picked up that idea from some speculative post. The only official information that I have is a mention that, on average, the servers are sending out about 7 copies of each work unit, though they get fewer results back. That would fit in with trying to get 5 identical results returned for each work unit, assuming that some unattended installations just keep haplessly requesting new work units without completing them while some computers have faulty FPUs. But I do not know.

Sorry I have no definitive information about this.
[Jul 28, 2005 10:58:48 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: Average process power? Gflops??

Hi musicmann,

Yesterday the WCG posted 12,850,066 points thats 535419 points per hour.

My PC scores 29.4 points per hour and benchmarks at 3.447 GFlops

So every hour there are 18211 PC's like mine crunching through Work Units for the WCG which would make 62775GFlops of raw number crunching power or 62.775 TeraFlops

This is the raw power of the WCG at the moment and would put is in 3rd place on the list of the top 500 supercomputer as of June this year.

There is redundancy however built into the system so that Work Units can be verified against each other due to the distributed nature of the Grid and the medical science that is being carried out here. I believe that each Work Unit is sent out about 5 times.

So the actual like for like comparison with a true Supercomputer would be around 12.5 TeraFlops give or take a GFlop or 2 which would still put the WCG a P14 in the list of the fastest supercomputers just 8 1/2 Months after the projects launch.

Which I think is just cool

take care and happy crunching

Regards
Dave


Hmmmmm nice to know. Thanks for reply.
[Jul 29, 2005 8:06:02 AM]   Link   Report threatening or abusive post: please login first  Go to top 
[ Jump to Last Post ]
Post new Thread