Core i7 running Seti

Message boards : Number crunching : Core i7 running Seti
Message board moderation

To post messages, you must log in.

Previous · 1 . . . 5 · 6 · 7 · 8 · 9 · 10 · 11 · Next

AuthorMessage
Profile zoom3+1=4
Volunteer tester
Avatar

Send message
Joined: 30 Nov 03
Posts: 65801
Credit: 55,293,173
RAC: 49
United States
Message 832076 - Posted: 19 Nov 2008, 7:55:49 UTC - in response to Message 832065.  
Last modified: 19 Nov 2008, 8:11:16 UTC

that's almost it...

However, here's what I had in mind ;)





As long as You've got the money, fine, Otherwise Xeon doesn't have the Memory controller onboard and Xeon is like LGA775 in that respect, i7 is the Intel equivalent of the Athlon, But with Hyperthreading and lots of speed. In any case i7 is not a dual quad core cpu socket design, "Who?" Could You explain this as I think I'm almost in over My head on this.


see that's why i posted the link to the host at rosetta... it seems to me that it's one of the new xeons based on the i7... ;)


Here's an entry on i7 right Here.

Intel i7 Wiki wrote:

Only one QuickPath interface: not intended for multi-processor motherboards.


You might be interested in Westmere which has enough QPI links for a DP motherboard. Beckton (Nehalem-EX) H2 2009 will be MP w/4 QPI Links(LGA1567, Not LGA1366), i7 currently is Bloomfield and isn't Capable of being a Xeon type cpu like You envision.
The T1 Trust, PRR T1 Class 4-4-4-4 #5550, 1 of America's First HST's
ID: 832076 · Report as offensive
Profile Crunch3r
Volunteer tester
Avatar

Send message
Joined: 15 Apr 99
Posts: 1546
Credit: 3,438,823
RAC: 0
Germany
Message 832140 - Posted: 19 Nov 2008, 15:55:18 UTC - in response to Message 832076.  


...
You might be interested in Westmere which has enough QPI links for a DP motherboard. Beckton (Nehalem-EX) H2 2009 will be MP w/4 QPI Links(LGA1567, Not LGA1366), i7 currently is Bloomfield and isn't Capable of being a Xeon type cpu like You envision.


So it appears to be a Gainestown then ?





Join BOINC United now!
ID: 832140 · Report as offensive
Profile zoom3+1=4
Volunteer tester
Avatar

Send message
Joined: 30 Nov 03
Posts: 65801
Credit: 55,293,173
RAC: 49
United States
Message 832149 - Posted: 19 Nov 2008, 16:32:42 UTC - in response to Message 832140.  
Last modified: 19 Nov 2008, 16:34:15 UTC


...
You might be interested in Westmere which has enough QPI links for a DP motherboard. Beckton (Nehalem-EX) H2 2009 will be MP w/4 QPI Links(LGA1567, Not LGA1366), i7 currently is Bloomfield and isn't Capable of being a Xeon type cpu like You envision.


So it appears to be a Gainestown then ?




According to the earlier Road Map You might want this one, Maybe, It's Your choice.

http://en.wikipedia.org/wiki/Xeon#Beckton

In any case, I'm not really sure here as I've seen no Intel reference motherboards.
The T1 Trust, PRR T1 Class 4-4-4-4 #5550, 1 of America's First HST's
ID: 832149 · Report as offensive
Profile Crunch3r
Volunteer tester
Avatar

Send message
Joined: 15 Apr 99
Posts: 1546
Credit: 3,438,823
RAC: 0
Germany
Message 832154 - Posted: 19 Nov 2008, 16:52:41 UTC - in response to Message 832149.  

That's all cool stuff you post there, however that all does not answer my question about what is/could be that host.

http://boinc.bakerlab.org/rosetta/show_host_detail.php?hostid=935401


Join BOINC United now!
ID: 832154 · Report as offensive
Profile jason_gee
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 24 Nov 06
Posts: 7489
Credit: 91,093,184
RAC: 0
Australia
Message 832167 - Posted: 19 Nov 2008, 17:59:53 UTC
Last modified: 19 Nov 2008, 18:00:35 UTC

[Intel64 Family 6 Model 26 Stepping 4]

Looks like a dual socket Nehalem based Xeon flavour to me ( X5570 maybe )

Maximum PC short mention of expected server CPU releases Q1 09.
"Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions.
ID: 832167 · Report as offensive
Mike Davis
Volunteer tester

Send message
Joined: 17 May 99
Posts: 240
Credit: 5,402,361
RAC: 0
Isle of Man
Message 832190 - Posted: 19 Nov 2008, 20:15:31 UTC - in response to Message 832154.  

That's all cool stuff you post there, however that all does not answer my question about what is/could be that host.

http://boinc.bakerlab.org/rosetta/show_host_detail.php?hostid=935401


Well... Seeing as his team is Intel Corp I guess it must be someone checking out how well a 'New Xeon' can go!
ID: 832190 · Report as offensive
Ivailo Bonev
Volunteer tester
Avatar

Send message
Joined: 26 Jun 00
Posts: 247
Credit: 35,864,461
RAC: 2
Bulgaria
Message 832201 - Posted: 19 Nov 2008, 20:52:42 UTC

I think Crunch3r is right, with 12GB RAM and 16 cores, with that strange "Intel(R) Xeon(R) CPU X55700" name...
Something like "Intel Bluff Creek" for me...
ID: 832201 · Report as offensive
Profile -= Vyper =-
Volunteer tester
Avatar

Send message
Joined: 5 Sep 99
Posts: 1652
Credit: 1,065,191,981
RAC: 2,537
Sweden
Message 832202 - Posted: 19 Nov 2008, 20:57:48 UTC

Well the re-build went like a charm..

Exchanged a Q9450 + P5E + 2x2GB DDR2 1040 Mhz ram to: i7 920 + P6T Deluxe + 2x3GB DDR3 1666 Mhz ram

Clocking it to 3.48 Ghz with the shipped original i7 cooler and it will be exchanged within two weeks..

DDR3 1333 Mhz Corsair modules clocked to 1666 without a hitch to, supplied a dominator cooler on top of them..

Impressive indeed...

Gonna follow up on the RAC later on..

Kind Regards Vyper

_________________________________________________________________________
Addicted to SETI crunching!
Founder of GPU Users Group
ID: 832202 · Report as offensive
Sirius B Project Donor
Volunteer tester
Avatar

Send message
Joined: 26 Dec 00
Posts: 24881
Credit: 3,081,182
RAC: 7
Ireland
Message 832207 - Posted: 19 Nov 2008, 21:08:10 UTC - in response to Message 832202.  

Well the re-build went like a charm..

Exchanged a Q9450 + P5E + 2x2GB DDR2 1040 Mhz ram to: i7 920 + P6T Deluxe + 2x3GB DDR3 1666 Mhz ram

Clocking it to 3.48 Ghz with the shipped original i7 cooler and it will be exchanged within two weeks..

DDR3 1333 Mhz Corsair modules clocked to 1666 without a hitch to, supplied a dominator cooler on top of them..

Impressive indeed...

Gonna follow up on the RAC later on..

Kind Regards Vyper



Hi Vyper, I'm not up to speed with Intel or the new boards/cpu's, but wouldn't it have been better to go for 3x 2gb to take advantage of the triple channel?
ID: 832207 · Report as offensive
archae86

Send message
Joined: 31 Aug 99
Posts: 909
Credit: 1,582,816
RAC: 0
United States
Message 832222 - Posted: 19 Nov 2008, 21:57:29 UTC - in response to Message 832190.  

Well... Seeing as his team is Intel Corp I guess it must be someone checking out how well a 'New Xeon' can go!
xPod has seven Nehalem-generation hosts, at least judging by "Family 6 Model 26 Stepping 4" designations.

All seem to report 16 CPUs to BOINC. The variety of OSs, the obviously erratic run times, and what seems to be the pre-release character of some or all of the parts (even if some are the Nehalem models released about November 16, they seem to have started before then) suggests this is some sort of sanity testing on multiple platforms, rather than a performance nut who somehow got his hands on early parts.

The same user has a similar array of hosts appearing on Einstein, Rosetta, Climate Prediction ... I did not sanity check, but as the usual number of "Family 6 Model 26" matches was 8, I suspect the same hosts with the same variety of OS etc. are visiting lots of the BOINC projects as yet one more form of torture testing. This would help to explain the low RACs.

ID: 832222 · Report as offensive
Profile -= Vyper =-
Volunteer tester
Avatar

Send message
Joined: 5 Sep 99
Posts: 1652
Credit: 1,065,191,981
RAC: 2,537
Sweden
Message 832237 - Posted: 19 Nov 2008, 23:11:35 UTC - in response to Message 832207.  

Hi Vyper, I'm not up to speed with Intel or the new boards/cpu's, but wouldn't it have been better to go for 3x 2gb to take advantage of the triple channel?


Heh , i meant the other way around ofc..

Trichannel it is.. :)

Kind reg. Vyper

_________________________________________________________________________
Addicted to SETI crunching!
Founder of GPU Users Group
ID: 832237 · Report as offensive
Profile -= Vyper =-
Volunteer tester
Avatar

Send message
Joined: 5 Sep 99
Posts: 1652
Credit: 1,065,191,981
RAC: 2,537
Sweden
Message 832369 - Posted: 20 Nov 2008, 7:08:24 UTC

Just curious if someone has even tested the i7 with HT disabled just to see the difference with "only" four cores?

Kind regards Vyper

_________________________________________________________________________
Addicted to SETI crunching!
Founder of GPU Users Group
ID: 832369 · Report as offensive
Profile zoom3+1=4
Volunteer tester
Avatar

Send message
Joined: 30 Nov 03
Posts: 65801
Credit: 55,293,173
RAC: 49
United States
Message 832370 - Posted: 20 Nov 2008, 7:18:08 UTC - in response to Message 832207.  

Well the re-build went like a charm..

Exchanged a Q9450 + P5E + 2x2GB DDR2 1040 MHz ram to: i7 920 + P6T Deluxe + 2x3GB DDR3 1666 MHz ram

Clocking it to 3.48 GHz with the shipped original i7 cooler and it will be exchanged within two weeks..

DDR3 1333 MHz Corsair modules clocked to 1666 without a hitch to, supplied a dominator cooler on top of them..

Impressive indeed...

Gonna follow up on the RAC later on..

Kind Regards Vyper



Hi Vyper, I'm not up to speed with Intel or the new boards/cpus, but wouldn't it have been better to go for 3x 2GB to take advantage of the triple channel?

For a cruncher? I'd think 3GB or 3x 1GB would do and six slots would be a waste of space and circuitry, That's why I've dropped any and all plans for Asus, MSI, DFI or Gigabyte, I'm looking at the Foxconn BloodRage instead as It seems to focused at overclockers. Oh and no Corsair for Me, I like Patriot better due to experience with both brands of ram. I may have to sell an Abit AW9D-MAX motherboard and a Q6600 B3 cpu of course. :D
The T1 Trust, PRR T1 Class 4-4-4-4 #5550, 1 of America's First HST's
ID: 832370 · Report as offensive
OzzFan Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 9 Apr 02
Posts: 15691
Credit: 84,761,841
RAC: 28
United States
Message 832580 - Posted: 20 Nov 2008, 23:20:38 UTC

I was looking at the Intel Core i7 940 and the Core i7 Extreme Edition (they're still using that name!?) 965 and I noticed that the 940 runs at 2.93GHz with a 4.8GT/s QPI while the 965 runs at 3.2GHz with a 6.4GT/s QPI. Is there a major performance loss with the 940 running at a slightly slower clock speed and a slower QPI speed?
ID: 832580 · Report as offensive
Profile SATAN
Avatar

Send message
Joined: 27 Aug 06
Posts: 835
Credit: 2,129,006
RAC: 0
United Kingdom
Message 832585 - Posted: 20 Nov 2008, 23:31:49 UTC

It might not be relevant to the new i7 chips. But once in a while I let the machine crunch on 8 cores at once and there is a clear drop in performance per unit. It actually crunches a single unit faster when only using four or less cores.

Who, has this been solved in the i7 chips?
ID: 832585 · Report as offensive
Olli

Send message
Joined: 25 Apr 07
Posts: 143
Credit: 2,089,162
RAC: 0
Finland
Message 832696 - Posted: 21 Nov 2008, 4:00:26 UTC - in response to Message 832585.  
Last modified: 21 Nov 2008, 4:02:40 UTC

Happy with Athlon 64 3500+. Really, maybe. Wrong thread for me.
ID: 832696 · Report as offensive
Profile -= Vyper =-
Volunteer tester
Avatar

Send message
Joined: 5 Sep 99
Posts: 1652
Credit: 1,065,191,981
RAC: 2,537
Sweden
Message 832736 - Posted: 21 Nov 2008, 6:54:01 UTC - in response to Message 832585.  

It might not be relevant to the new i7 chips. But once in a while I let the machine crunch on 8 cores at once and there is a clear drop in performance per unit. It actually crunches a single unit faster when only using four or less cores.

Who, has this been solved in the i7 chips?


As long as you crunch more data and occupy the pipeline more then you will always get a "clear" drop in performance in time, especially how HT is designed.
But if you sum up how fast it churns out the data then you should see that you get more work done per day instead..

8 units at 60 minutes is a more productive machine than 4 units at 38 minutes..

Kind regards Vyper

_________________________________________________________________________
Addicted to SETI crunching!
Founder of GPU Users Group
ID: 832736 · Report as offensive
kittyman Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 9 Jul 00
Posts: 51469
Credit: 1,018,363,574
RAC: 1,004
United States
Message 832739 - Posted: 21 Nov 2008, 7:00:03 UTC - in response to Message 832736.  

It might not be relevant to the new i7 chips. But once in a while I let the machine crunch on 8 cores at once and there is a clear drop in performance per unit. It actually crunches a single unit faster when only using four or less cores.

Who, has this been solved in the i7 chips?


As long as you crunch more data and occupy the pipeline more then you will always get a "clear" drop in performance in time, especially how HT is designed.
But if you sum up how fast it churns out the data then you should see that you get more work done per day instead..

8 units at 60 minutes is a more productive machine than 4 units at 38 minutes..

Kind regards Vyper

I am sure you are correct........total output should increase....but it will not scale at 2:1 because you are splitting total resources available between 8 threads rather than devoting all resources to the support of 4 threads......
"Freedom is just Chaos, with better lighting." Alan Dean Foster

ID: 832739 · Report as offensive
Profile -= Vyper =-
Volunteer tester
Avatar

Send message
Joined: 5 Sep 99
Posts: 1652
Credit: 1,065,191,981
RAC: 2,537
Sweden
Message 832742 - Posted: 21 Nov 2008, 7:05:36 UTC - in response to Message 832739.  

I am sure you are correct........total output should increase....but it will not scale at 2:1 because you are splitting total resources available between 8 threads rather than devoting all resources to the support of 4 threads......


That is another way to describe same thing that i wrote but in another point of view.. :)

Happy crunching..

Kind Regards Vyper

_________________________________________________________________________
Addicted to SETI crunching!
Founder of GPU Users Group
ID: 832742 · Report as offensive
kittyman Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 9 Jul 00
Posts: 51469
Credit: 1,018,363,574
RAC: 1,004
United States
Message 832745 - Posted: 21 Nov 2008, 7:10:07 UTC - in response to Message 832742.  

I am sure you are correct........total output should increase....but it will not scale at 2:1 because you are splitting total resources available between 8 threads rather than devoting all resources to the support of 4 threads......


That is another way to describe same thing that i wrote but in another point of view.. :)

Happy crunching..

Kind Regards Vyper

Soon............meowwwwwwwwwwwwwwwwwwwwwwwwwwwwwwww.
"Freedom is just Chaos, with better lighting." Alan Dean Foster

ID: 832745 · Report as offensive
Previous · 1 . . . 5 · 6 · 7 · 8 · 9 · 10 · 11 · Next

Message boards : Number crunching : Core i7 running Seti


 
©2024 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.