Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
![]() |
World Community Grid Forums
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
No member browsing this thread |
Thread Status: Active Total posts in this thread: 41
|
![]() |
Author |
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Does having an overclock have a benefit? I can't really tell because each Work Unit has a different time span and it seems the estimated time doens't change when I toggle my CPU speed.
I have an E6600 @ 3.2 GHz. Stable. Here's what the client reports. 11/14/2007 2:59:12 PM||Starting BOINC client version 5.10.22 for windows_intelx86 11/14/2007 2:59:12 PM||log flags: task, file_xfer, sched_ops 11/14/2007 2:59:12 PM||Libraries: libcurl/7.16.1 OpenSSL/0.9.8e zlib/1.2.3 11/14/2007 2:59:12 PM||Data directory: C:\Program Files\BOINC 11/14/2007 2:59:12 PM||Processor: 2 GenuineIntel Intel(R) Core(TM)2 CPU 6600 @ 2.40GHz [x86 Family 6 Model 15 Stepping 6] 11/14/2007 2:59:12 PM||Processor features: fpu tsc pae nx sse sse2 mmx 11/14/2007 2:59:12 PM||OS: Microsoft Windows XP: Professional Edition, Service Pack 2, (05.01.2600.00) 11/14/2007 2:59:12 PM||Memory: 2.00 GB physical, 3.85 GB virtual 11/14/2007 2:59:12 PM||Disk: 78.13 GB total, 66.58 GB free 11/14/2007 2:59:12 PM||Local time is UTC -8 hours 11/14/2007 2:59:12 PM|World Community Grid|URL: http://www.worldcommunitygrid.org/; Computer ID: 323671; location: (none); project prefs: default 11/14/2007 2:59:12 PM||General prefs: from World Community Grid (last modified 2007-10-05 22:41:15) 11/14/2007 2:59:12 PM||Host location: none 11/14/2007 2:59:12 PM||General prefs: using your defaults 11/14/2007 2:59:12 PM||Preferences limit memory usage when active to 2046.48MB 11/14/2007 2:59:12 PM||Preferences limit memory usage when idle to 2046.48MB 11/14/2007 2:59:12 PM||Preferences limit disk usage to 3.73GB So is the crunching process going any faster? |
||
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Please don't overclock. Overclocking substantially increases the risk of invalid results.
|
||
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Out of my ~400 units for the last month, only two of them were invalid. If my CPU is able to crunch through more work units more quickly and efficiently, there will be more benefit towards the study.
----------------------------------------My CPU is overclocked and stable for other purposes other than crunching. [Edit 1 times, last edit by Former Member at Nov 15, 2007 7:33:35 AM] |
||
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
Blizzie, my team houses on a website called "A Gathering of Tweakers". Many OC, but you're really on your own in terms of having to watch the results to remain valid. What it does to the CPU and other components if not cooled properly nor on the lifespan I don't know.... it's an activity i wont partake in now.
----------------------------------------As such, the sciences were designed to run on 'stock' speed machines. OCing can change timings of different steps in the science that need to converge. An invalid occurs if even a single bit went the other way or arrived too late! As for flipping the OC switch, BOINC adjusts only upon learning from current historic progress, it will thus take time like a day or more to get the predicted Edit: strike and substitute.
WCG
----------------------------------------Please help to make the Forums an enjoyable experience for All! [Edit 1 times, last edit by Sekerob at Nov 15, 2007 2:21:26 PM] |
||
|
crooks_uk
Veteran Cruncher England Joined: Nov 25, 2004 Post Count: 1011 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Out of my ~400 units for the last month, only two of them were invalid. If my CPU is able to crunch through more work units more quickly and efficiently, there will be more benefit towards the study. My CPU is overclocked and stable for other purposes other than crunching. Blizzie, I agree with the above. All of my machines have some degree of overclocking applied to them and I have never had any issues. If you know what you are doing and are using quality components, I do not see any issues. Got to be honest though, I never overclock by more than 15% (10% on some machines), purely do to the fact that my processors run at 100% 24/7/365.
Be a part of the largest UK team:
http://www.worldcommunitygrid.org/team/viewTeamInfo.do?teamId=L721SPD4BN1 |
||
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Nothing wrong with overclocking as long as you know what you are doing and they are stable, but as Sek says you need to keep an eye out for errors.
Like crooks_uk, all my machines are overclocked to some degree, without any problems, and the benefits in getting more work done is obvious. To tell those from teams such as XtremeSystems not to overclock would cause mass panic ![]() |
||
|
jal2
Senior Cruncher USA Joined: Apr 28, 2007 Post Count: 422 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
my brief attempt at OC showed that some projects are more sensitive than others. If I remember correctly, FAAH went ballistic on one of my OC machines, while the other projects just finished faster. ACH had one invalid on another OC machine, but that WU had 8 invalid, 1 error, and 13 valid hits, so I question whether that was just a WU glitch.
----------------------------------------I also found it interesting that the floating point unit showed a significantly higher boost than the integer unit for the specified OC. I would have expected a more linear boost, but that's a discussion for another thread. ![]() |
||
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Alright. I'm going to stick with what I have now and see if there's an increase in my progress speed. I doubt that they'll be invalid units but I'll keep an eye out for them.
|
||
|
twilyth
Master Cruncher US Joined: Mar 30, 2007 Post Count: 2130 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
You should probably post this in the WCG team section of XtrememSystems.org/forum. You'll need to register first of course - but that's just for the forum not for the WCG team.
----------------------------------------I have 2 Q6600's "designed" to run at 2.4 ghz. I'm running one at 3.6ghz and it has only produced errors when I had changed something and it was not yet stable. The other is at 3.2ghz and I won't bump that up again until the thermal paste has settled. There's no issue with running apps at ANY speed that the chip will tolerate. The only difference for example between chips with the same architecture, stepping, etc that run at different rated speeds is that the faster chips can get away with it and the slower ones can't. There's no design limitation on speed until you get into truly extreme overclocking. Using liquid nitrogen I believe, the Q6600 has been run at 5ghz. Will every q6600 be able to do that? No. But some can. The best way to check stability and temperatures is to run something like Prime95 (select options --> torture test --> 'in-place large ffts') - dl from http://www.mersenne.org/freesoft.htm (thanks to XS_Olympic). If you can run this for a couple hours and not get any errors or thermal checks, you will probably be ok. But remember, if you are pushing the envelope, your system will be much more sensitive to external factors like ambient temperature, dust build up in heat sinks, etc. If you don't keep an eye on your system daily and check it's health status with something like speedfan or a similar utility, you could fry your system. There is a software package called the oc bible that has a lot of free utilities for overclockers. speedfan - http://www.almico.com/sfdownload.php oc bible - http://www.softpedia.com/get/System/System-Miscellaneous/OcBible.shtml ![]() ![]() |
||
|
mike047
Senior Cruncher Joined: Aug 22, 2006 Post Count: 262 Status: Offline Project Badges: ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Please don't overclock. Overclocking substantially increases the risk of invalid results. Can you direct me to conclusive information to support this statement. Unstable equipment will not be reliable, but myself and MANY others ![]() With the exception of this MiniMac I am currently on, ALL 46 of my boxes work on a 24/7 overclock[20-35%].
mike
Crunch Hard, Crunch Often ![]() |
||
|
|
![]() |