-bigadv benchmark machine

The most demanding Projects are only available to a small percentage of very high-end servers.

Moderators: Site Moderators, PandeGroup

-bigadv benchmark machine

Postby PinHead » Mon Feb 13, 2012 4:16 am

Has a -bigadv benchmark machine been announced? And I just can't seem to locate it!

I've read a bunch of FAQ's and thread's that just take me back in time to speculation about what hardware may or may not make the grade.

Somebody "link" me up.
PinHead
 
Posts: 352
Joined: Tue Jan 24, 2012 3:43 am

Re: -bigadv benchmark machine

Postby Grandpa_01 » Mon Feb 13, 2012 4:58 am

This is as close as you are going to get. https://docs.google.com/spreadsheet/ccc ... ERFE#gid=0
Image
4 - SuperMicro H8QGi-F AMD 6xxx=192 cores @ 2.8Ghz
3 - SM X9QRI-f+ Intel 4650 = 192 cores/threads @ 3Ghz
2 - I7 980X 4.4Ghz 2 - GTX680
1 - 2700k 4.4Ghz GTX680
Total = 419 cores folding
User avatar
Grandpa_01
 
Posts: 1747
Joined: Wed Mar 04, 2009 7:36 am

Re: -bigadv benchmark machine

Postby bruce » Mon Feb 13, 2012 5:03 am

Grandpa_01 wrote:This is as close as you are going to get. https://docs.google.com/spreadsheet/ccc ... ERFE#gid=0


That is ABSOLUTELY not the announcement of a benchmark machine. The points are still being realigned, Any collection of obsolete data is just that -- obsolete data.

Dr kasson has said clearly that the final plan has not been established yet. Results of any kind of early testing is always subject to change.
bruce
Site Admin
 
Posts: 16849
Joined: Thu Nov 29, 2007 10:13 pm
Location: So. Cal.

Re: -bigadv benchmark machine

Postby Flathead74 » Mon Feb 13, 2012 5:06 am

The only -bigadv benchmark machine description I've seen mentioned is located here:

http://folding.stanford.edu/English/FAQ-Points#ntoc11
Flathead74
 
Posts: 618
Joined: Sun Dec 02, 2007 6:08 pm
Location: Central New York

Re: -bigadv benchmark machine

Postby PinHead » Mon Feb 13, 2012 5:28 am

Flathead74 wrote:The only -bigadv benchmark machine description I've seen mentioned is located here:

http://folding.stanford.edu/English/FAQ-Points#ntoc11


With all due respect, I think that is 'old' data:

•Primary CPU Intel Xeon UP Quad Core X3450 2.66Ghz 8Mb Cache 95w

As the new specs require 16 cores.

Maybe a bench mark WU in a given time would be a better spec? Since a current 16 core chip doesn't seem to cut the mustard.
PinHead
 
Posts: 352
Joined: Tue Jan 24, 2012 3:43 am

Re: -bigadv benchmark machine

Postby 7im » Mon Feb 13, 2012 6:34 am

The latest published config is a 2.6 GHz i5.
Please do not mistake my brevity as dispassion or condescension. I recognize the time you spend reading the forum is time you could use elsewhere, so my short responses save you time. Please do not hesitate to ask for clarification if I was too terse.
User avatar
7im
 
Posts: 13303
Joined: Thu Nov 29, 2007 4:30 pm
Location: Arizona

Re: -bigadv benchmark machine

Postby 7im » Mon Feb 13, 2012 6:38 am

Grandpa_01 wrote:This is as close as you are going to get. https://docs.google.com/spreadsheet/ccc ... ERFE#gid=0



Why publish the final deadline in a chart when the points drop off at the preferred deadline.

PD on a p8101 is 2.4 days, FD is 4.0 days.
User avatar
7im
 
Posts: 13303
Joined: Thu Nov 29, 2007 4:30 pm
Location: Arizona

Re: -bigadv benchmark machine

Postby Patriot » Tue Feb 21, 2012 12:01 am

7im wrote:
Grandpa_01 wrote:This is as close as you are going to get. https://docs.google.com/spreadsheet/ccc ... ERFE#gid=0



Why publish the final deadline in a chart when the points drop off at the preferred deadline.

PD on a p8101 is 2.4 days, FD is 4.0 days.


Because its a spreadsheet, and the values are calculated off final not preferred.
Image
Patriot
 
Posts: 90
Joined: Sat Aug 16, 2008 2:04 pm

Re: -bigadv benchmark machine

Postby PinHead » Tue Feb 21, 2012 4:58 am

7im wrote:The latest published config is a 2.6 GHz i5.


So can someone publish the 2.6GHz i5 on a 6904 tpf?

Need to benchmark a 16 core 2.1 GHz vs the i5, to see if it is 1.5X.

Thanks,
pH
PinHead
 
Posts: 352
Joined: Tue Jan 24, 2012 3:43 am

Re: -bigadv benchmark machine

Postby 7im » Tue Feb 21, 2012 5:05 am

The i5 ppd is published in the FAQ for smp. ~6k Pppd. Work the math back from there.
User avatar
7im
 
Posts: 13303
Joined: Thu Nov 29, 2007 4:30 pm
Location: Arizona

Re: -bigadv benchmark machine

Postby PinHead » Tue Feb 21, 2012 5:16 am

7im wrote:The i5 ppd is published in the FAQ for smp. ~6k Pppd. Work the math back from there.


hmmm,

Since an i5 won't do -bigadv, guess I'll do some extrapolation!
PinHead
 
Posts: 352
Joined: Tue Jan 24, 2012 3:43 am

Re: -bigadv benchmark machine

Postby Patriot » Tue Apr 24, 2012 6:10 am

While not an official benchmark machine...
This spreadsheet includes more data points than they make decisions off of anyways.

I updated the spreadsheet to properly cut off bonus if preferred deadline is not met.
I also sanity checked the data again and found it to be accurate.
https://docs.google.com/spreadsheet/ccc ... zBnRklERFE
Patriot
 
Posts: 90
Joined: Sat Aug 16, 2008 2:04 pm

Re: -bigadv benchmark machine

Postby DocJonz » Wed May 02, 2012 6:26 pm

Can I clarify what the current -bigadv requirement actually is?

- Is it a 16 'physical' core capability? or 16 'threads'?
- What is the memory requirement? - XXGB per core (or per thread) of a toal of YYGB?
- Do they run under both Windows and Linux?
User avatar
DocJonz
 
Posts: 296
Joined: Thu Dec 06, 2007 6:31 pm
Location: United Kingdom

Re: -bigadv benchmark machine

Postby bollix47 » Wed May 02, 2012 6:39 pm

The P8101 WUs probably won't finish within the 2.4 days preferred deadline on a 16 thread CPU unless it's highly overclocked and even then it might be close. My 8c/16t dual xeon system running at 2.54 Ghz is not capable(not even close) of making that deadline and has been doing regular SMP projects since the 8101 was released.

Memory of ~100 MB per core is being used on P8101 that's currently running on my 12c/24t computer.

AFAIK just Linux for now.
Last edited by bollix47 on Sun May 06, 2012 5:12 pm, edited 3 times in total.
bollix47
Site Moderator
 
Posts: 2807
Joined: Sun Dec 02, 2007 5:04 am
Location: Canada

Re: -bigadv benchmark machine

Postby Macaholic » Wed May 02, 2012 6:52 pm

Dr. Pande posted on his blog that 16 Cores are required. You can also peruse, this thread, at your convenience. And the thread here will give you some historical perspective as to what to expect.
Fold! It does a body good!™
User avatar
Macaholic
Super Moderator
 
Posts: 903
Joined: Thu Nov 29, 2007 11:57 pm
Location: 1 Infinite Loop

Next

Return to SMP with bigadv

Who is online

Users browsing this forum: No registered users and 1 guest