| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: Help Conquer Cancer
Thread: Updated Windows Help Conquer Cancer Application (v 6.42) |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 52
|
|
| Author |
|
|
Falconet
Master Cruncher Portugal Joined: Mar 9, 2009 Post Count: 3295 Status: Offline Project Badges: 
|
The compiler updates just took my Pentium computer out of the running. Oh well, I still have four more computers doing work. What do you mean? AMD Ryzen 5 1600AF 6C/12T 3.2 GHz - 85W AMD Ryzen 5 2500U 4C/8T 2.0 GHz - 28W AMD Ryzen 7 7730U 8C/16T 3.0 GHz |
||
|
|
John Doze
Cruncher Joined: Dec 16, 2005 Post Count: 1 Status: Offline Project Badges:
|
I cannot run HCC v 6.42 on my system. I have a quad processor, 4Gb of ram and the OS is windows7 64bit, All other WU run fine with no problems. But HCC fails with computation errors after only a few seconds. Has any one else have this problem or know of a fix.
----------------------------------------[Edit 1 times, last edit by John Doze at Apr 21, 2011 8:51:32 AM] |
||
|
|
Somervillejudson@netscape.net
Veteran Cruncher USA Joined: May 16, 2008 Post Count: 1065 Status: Offline Project Badges: 
|
Yes have a few computational errors! Now run time down to slightly over one hour per WU! Only running HCC on all machines which is about fifty four computational days per actual day.
|
||
|
|
z2000
Advanced Cruncher Joined: Feb 27, 2011 Post Count: 116 Status: Offline |
Since the Linux version cannot get a faster version does this mean that the Linux version is still faster than HCC 6.42 for Windows?
---------------------------------------- |
||
|
|
sk..
Master Cruncher http://s17.rimg.info/ccb5d62bd3e856cc0d1df9b0ee2f7f6a.gif Joined: Mar 22, 2007 Post Count: 2324 Status: Offline Project Badges:
|
Why are they taking so long on a i7 920? Do you have BOINC set to "Use at most 100% of CPU time" and is it set to run "While in use"?I would have thought I would be getting much more from an SB 3.5GHz (HT on); 1.28h 2003 x64 At the present time it looks like a highly overclocked IC2Q matches an i7-920, and an SB is only about 12% faster overall. Perhaps there is still room for app improvement. Will there be an x64 app for HCC? |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
eting average of arround 1.40 on my i7960 hyperthreading on..
but my little laptop with only 2.8 is avg arround 1.15 (dual processor |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
The HCC Crunching charts shows a project average run time change from about 2.41 to about avg. 1.60 hours per task or about 50% speed gain, which suggest that the Linux version is still outperforming the new Windows v 6.42
See chart: http://www.worldcommunitygrid.org/forums/wcg/viewthread_thread,18662 --//-- |
||
|
|
nanoprobe
Master Cruncher Classified Joined: Aug 29, 2008 Post Count: 2998 Status: Offline Project Badges: 
|
The HCC Crunching charts shows a project average run time change from about 2.41 to about avg. 1.60 hours per task or about 50% speed gain, which suggest that the Linux version is still outperforming the new Windows v 6.42 It is. 
In 1969 I took an oath to defend and protect the U S Constitution against all enemies, both foreign and Domestic. There was no expiration date.
  |
||
|
|
Ingleside
Veteran Cruncher Norway Joined: Nov 19, 2005 Post Count: 974 Status: Offline Project Badges: 
|
The HCC Crunching charts shows a project average run time change from about 2.41 to about avg. 1.60 hours per task or about 50% speed gain, which suggest that the Linux version is still outperforming the new Windows v 6.42 See chart: http://www.worldcommunitygrid.org/forums/wcg/viewthread_thread,18662 --//-- Hmm, maybe I'm overlooking something, but how can you say anything about the speed under Linux and Windows based on the average run-time?  Having some "fun with numbers"...  Old application: Let's say Linux/Mac used 1.61 hours/task. Let's also say Windows used 2.7 hours/task. Meaning, Windows used 1.68x longer per task. A quick look on BoincStats reveals that for WCG, Linux/Mac has 17.68% of the total RAC. (17.68% * 1.68)/(17.68% * 1.68 + (100% - 17.68%)) = 26.5% So, based on Linux/Mac having higher speed before, this indicates that Linux/Mac crunched 26.5% of the HCC-tasks, and Windows crunched 73.5% of the HCC-tasks before. This gives the following average speed: 1.61 hours/task * 0.265 + 2.7 hours/task 0.735 = 2.41 hours/task with old application. New application: Linux/Mac still uses 1.61 hours/task. Windows uses 1.61 hours/task. Meaning, average is now... 1.61 hours/task.  Well, both 1.61 hours and 2.7 hours was taken out of thin air, let's look on some real numbers under windows: i7: old average 9761.9 s/task, max 11054.4 - min 8833.3 - 25% variation, 301 results. i7: New average, 6698.3 s/task, old was 1.46x longer, max - 7487.5 - min 6081.4 - 23% variation, 37 results. Amd: old average 8366.5 s/task, new 5004.7 s/task, meaning old was 1.67x longer. Variation, old 14%, new 44% with possibly one outlier, after removing this outlier the variation was still... 26% with 52 results.  "I make so many mistakes. But then just think of all the mistakes I don't make, although I might." [Edit 1 times, last edit by Ingleside at Apr 26, 2011 10:01:00 PM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Gained a lot of speed. Went from a average of 2.4 to 1.1 hours.
On Intel Duo E8400 @3.00 Ghz with XP x86. Not bad. Crunching on.  |
||
|
|
|