I'm still under 24 cores

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

Moderators: Site Moderators, PandeGroup

Re: I'm still under 24 cores

Postby bollix47 » Mon Jul 09, 2012 12:24 pm

Instead of just 12 cores have you considered using -smp 15(5x3). I have used it in the past and it does work. Since it's under 16 you should get the 690x projects.
bollix47
 
Posts: 3493
Joined: Sun Dec 02, 2007 5:04 am
Location: Canada

Re: I'm still under 24 cores

Postby Grandpa_01 » Mon Jul 09, 2012 1:18 pm

No it will not work for getting the 69xx WU's, the amount of cores come from what the OS reports to fah client not what you have it set for. I have 48 cores per rig so 48 cores are reported to the server by fah. There are some gray ways around it but I am not using them, if Stanford wants me to shut them down and start up the 12 core rigs that is what I will do. Hopefully it is just an AS computer glitch and will get fixed.
Image
2 - SM H8QGi-F AMD 6xxx=112 cores @ 3.2 & 3.9Ghz
5 - SM X9QRI-f+ Intel 4650 = 320 cores @ 3.15Ghz
2 - I7 980X 4.4Ghz 2-GTX680
1 - 2700k 4.4Ghz GTX680
Total = 464 cores folding
User avatar
Grandpa_01
 
Posts: 1757
Joined: Wed Mar 04, 2009 7:36 am

Re: I'm still under 24 cores

Postby 7im » Mon Jul 09, 2012 1:50 pm

Stanford wants people to fold. How those people fold is completely up to each individual. There is no implied message directed at anyone.
How to provide enough information to get helpful support
Tell me and I forget. Teach me and I remember. Involve me and I learn.
User avatar
7im
 
Posts: 14648
Joined: Thu Nov 29, 2007 4:30 pm
Location: Arizona

Re: I'm still under 24 cores

Postby Jesse_V » Mon Jul 09, 2012 3:28 pm

7im wrote:Stanford wants people to fold. How those people fold is completely up to each individual. There is no implied message directed at anyone.

That's true, having a variety of folding clients is good. I think it comes down to points for some people. But you're still right, because IIRC Dr. Kasson invented the bigadv system, and he's not at Stanford right now is he? ;)
Pen tester at Cigital/Synopsys
User avatar
Jesse_V
 
Posts: 2773
Joined: Mon Jul 18, 2011 4:44 am
Location: USA

Re: I'm still under 24 cores

Postby Grandpa_01 » Mon Jul 09, 2012 5:52 pm

7im wrote:Stanford wants people to fold. How those people fold is completely up to each individual. There is no implied message directed at anyone.


Yes there is a message implied or otherwise the value placed on the 69xx is greater than the value of the 8101 per amount of work done and power and equipment require to do the given job thus Stanford places a greater value on the folding machines with less than 16 cores. Stanford gets X amount of donation from me if they say with the points system that y has a greater value than x then I will run y. The point system currently says y is more valuable so x will be shut down and y will run. That is Stanford’s choice not mine they are the ones that control the points system. :ewink:

It does not take a rocket scientist to see the implied message here. :lol:
User avatar
Grandpa_01
 
Posts: 1757
Joined: Wed Mar 04, 2009 7:36 am

Re: I'm still under 24 cores

Postby 7im » Mon Jul 09, 2012 6:40 pm

Correlation is not causation. Any non-rocket scientist knows that from junior high science class. And we both know there is more to the story here about 690x vs 810x. The point VALUES currently says y is higher than x. So yes, please do run y WUs until they are gone. The sooner the better. Then run something else. ;)

But that is YOUR interpretation of the existing circumstances of the publicly available information. You may read in to it whatever message you want. Pande Group does NOT actively sending out any specific message, other than on the NEWS BLOG. Servers break and get repaired, changing the WU mix all the time. And Projects end when they end. There is no implied message there.
User avatar
7im
 
Posts: 14648
Joined: Thu Nov 29, 2007 4:30 pm
Location: Arizona

Re: I'm still under 24 cores

Postby bruce » Mon Jul 09, 2012 6:48 pm

Jesse_V wrote:
7im wrote:Stanford wants people to fold. How those people fold is completely up to each individual. There is no implied message directed at anyone.

That's true, having a variety of folding clients is good. I think it comes down to points for some people. But you're still right, because IIRC Dr. Kasson invented the bigadv system, and he's not at Stanford right now is he? ;)


Loosely speaking, I sometimes use "Stanford" when I probably should use "Pande Group" or "Folding@home." The fact that the research started by Dr. Vijay Pande is spreading to a variety of locations doesn't really change how the Assignment Server policies are set or how the points are benchmarked. The fact that Dr. Kasson is not at Stanford riht now doesn't mean he's no longer part of the "Pande Group" or "Folding@home" (or "Stanford" if I happen to use that term inaccurately.) Every year i becomes a more widespread, world-wide consortium of scientists studying detailed atomic interactions of proteins.
bruce
 
Posts: 22623
Joined: Thu Nov 29, 2007 10:13 pm
Location: So. Cal.

Re: I'm still under 24 cores

Postby mike.million » Tue Jul 17, 2012 12:29 am

Grandpa_01 wrote:Yes a 4.3Ghz OC on a sr2 is a heavy OC the sr2's are not know as the best OCing rigs out there they can be stable for days then for no apparent reason go bonkers, they are known to loose there OC on reboot also for no apparent reason. You may be lucky and have 1 of the good ones but that is a pretty rare occurrence. I tried 2 different sr2's at 1 time I hated those board's with a Passion, they would run good for a while then just go crazy, very frustrating. :(


What the SR2 not know as a good OCer???? OMG, It IS the only rig you can overclock a Xeon on, so it is the best Xenon overclocker in the history of Xenon's!

My 2.6 ghz 5650 was a very average folder before it was oc'ed, don't remember the exact numbers but I was very dissapointed. Would you fold one not oc'ed?

Mine has been at 4 ghz for ~2 years now and yes twice while rebooting it lost my bios settings, and once it crashed when I plugged in a finger drive.

Man I oc'ed a bunch of boards over the years and this is the most solid, stable and constant board I can ever remember. Especially at this sort of overclock and usage.

From the BH6 with my Celeron 300a to my last I7 board, they have all done the same sort of things as my SR2, and often were much worse.

How many times have I had a highly oc'ed mb crash all the way back to the stone age? Where I had to reflash the bios, to get it to be more useful than a rock. Often the mb would never get close to the old oc after a bad crash, if it ever overclocked again.

I used mostly ASUS mb's over the years with a few Abit, Evga, and Gigabyte mbs in the mix.

They are all pretty much the same, when you run them way over the designed operating temps and bus speeds, they crash.

fold on
Image
mike.million
 
Posts: 63
Joined: Wed Jan 14, 2009 3:41 am

Re: I'm still under 24 cores

Postby Grandpa_01 » Tue Jul 17, 2012 3:19 am

Congrats you are one of the lucky few. But odds are you could take the same CPU and put it in another single processor board and get between 4.4Ghz and 4.5Ghz out of it. They have amassed quite a reputation for being unreliable if you try to push the cpu's to where they should be able to go to. Unfortunately they are pretty much the only choice Intel. Hopefully the will do better with the new evga duel processor board. The only thing we can do is hope. :D
User avatar
Grandpa_01
 
Posts: 1757
Joined: Wed Mar 04, 2009 7:36 am

Re: I'm still under 24 cores

Postby mike.million » Thu Jul 19, 2012 3:44 am

Well I never considered a single processor mb, so have no idea if any do oc nor do I care.

Only a dual or quad set up was is of any interest to me so eh, I will amend my statement to dual processor mb capable of overclocking a Xenon.

In any case it works great, at least I think a stable 1.33 ghz overclock is great on a 2.67 ghz processor. I firmly believe, if all the people that had a problem with an extreme oc, 4.3 or 4.4 or 4.5, would have been happy with 4 ghz, they never would have had a problem and a lot more points.

Then they seem surprised.

obtw I was shocked how much less heat my SR2 is generating, even at 4 ghz, when I turned off hyper threading (to increase my points by only having 12 cores). I mean shocked, this room has always been just bearable with my SR2 folding. Now a 6901 takes around a minute more per section, but my room is cool! First day I came home I thought my SR2 had stopped folding as the room was just too cool. Wish I had done this a long time ago.
mike.million
 
Posts: 63
Joined: Wed Jan 14, 2009 3:41 am

Re: I'm still under 24 cores

Postby bruce » Thu Jul 19, 2012 4:19 am

mike.million wrote:obtw I was shocked how much less heat my SR2 is generating, even at 4 ghz, when I turned off hyper threading (to increase my points by only having 12 cores). I mean shocked, this room has always been just bearable with my SR2 folding. Now a 6901 takes around a minute more per section, but my room is cool! First day I came home I thought my SR2 had stopped folding as the room was just too cool. Wish I had done this a long time ago.
(Yes, it's off-topic. No response is needed --- it's just an idle thought.)

I wonder if that also applies to other Intel systems. How much can I reduce my heat if I turn off HT on an i7 running SMP:8 and run SMP:4? In other words, maybe an i5 would have been a more prudent choice.

Something to test......
bruce
 
Posts: 22623
Joined: Thu Nov 29, 2007 10:13 pm
Location: So. Cal.

Re: I'm still under 24 cores

Postby ra_alfaomega » Fri Jul 20, 2012 10:49 am

I was curious about project 8101 and I tried just once to get one, using a tweak in ubuntu everybody knows about it( to see I have 16 cores). On a 3930k@ 4,4 Ghz with only 12 threads I managed to complete in time with a time/frame of ~32min. I can overclock beyond 4,6 so it's interesting how much power this cpu has with only 12 threads.Memory was just 1333mhz, maybe with 1866 or 2133 tpf will improve.
ra_alfaomega
 
Posts: 25
Joined: Tue Aug 23, 2011 9:00 am

Re: I'm still under 24 cores

Postby Grandpa_01 » Sat Jul 21, 2012 2:55 am

PG has said the core hack is acceptable as long as you are able to meet the preferred deadline. They have stated it is the speed the WU is completed in that is important not the core count. :ewink: If you do not mind could you post your results here. https://docs.google.com/spreadsheet/ccc ... DcHc#gid=0 So that we have an accurate data base of what will work and what will not work on the new bigadv.
User avatar
Grandpa_01
 
Posts: 1757
Joined: Wed Mar 04, 2009 7:36 am

Re: I'm still under 24 cores

Postby bruce » Sat Jul 21, 2012 4:34 am

Grandpa_01 wrote:PG has said the core hack is acceptable as long as you are able to meet the preferred deadline.
Please show me where the Pande Group said that. When the Pande Group says "Donors should not..." it DOES NOT mean the activity is "considered acceptable."

I'm sure there are a number of folks who don't actually look things up and soon come to believe everything their teammates say. For some, their sense of morality ends with something like "The PG is not preventing me from doing X, so it must be acceptable. On the other hand, why should scientists to be forced to waste their time forcing people to conform to good practices instead of paying attention to the science. Donors should be respectful enough of the scientific goals OF FAH to cooperate with the PG's recommendations, even if that's not their personal goal.

Consider this post and the referenced FAQs: http://folding.typepad.com/news/2011/05 ... s-faq.html
bruce
 
Posts: 22623
Joined: Thu Nov 29, 2007 10:13 pm
Location: So. Cal.

Re: I'm still under 24 cores

Postby Grandpa_01 » Sat Jul 21, 2012 4:37 am

bruce wrote:
Grandpa_01 wrote:PG has said the core hack is acceptable as long as you are able to meet the preferred deadline.
Please show me where the Pande Group said that. When the Pande Group says "Donors should not..." it DOES NOT mean the activity is "considered acceptable."

I'm sure there are a number of folks who don't actually look things up and soon come to believe everything their teammates say. For some, their sense of morality ends with something like "The PG is not preventing me from doing X, so it must be acceptable. On the other hand, why should scientists to be forced to waste their time forcing people to conform to good practices instead of paying attention to the science. Donors should be respectful enough of the scientific goals OF FAH to cooperate with the PG's recommendations, even if that's not their personal goal.

Consider this post and the referenced FAQs: http://folding.typepad.com/news/2011/05 ... s-faq.html


Here you go viewtopic.php?f=16&t=20036&start=30#p199180
User avatar
Grandpa_01
 
Posts: 1757
Joined: Wed Mar 04, 2009 7:36 am

PreviousNext

Return to SMP with bigadv

Who is online

Users browsing this forum: No registered users and 1 guest

cron