| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: Computing for Clean Water Forum
Thread: 64 Bit Computing for Clean Water application deployed |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 32
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
 My current impression is of a slight speedup, but less than 10% on my AMD quad-core. If people ask me, I will answer 'up to 10%' (ambiguous answer) until either the WCG staff or the project scientists give their figures for average difference in speed. Lawrence |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Day 3 I think of 64 bit availability and morning averages for the project give 4.91 hours, from before 5.01 i.e. 1/10th of an hour gain for the project or 2%
Last 7 days hourly average: 5.02501 14th 5.03419 15th 5.00696 16th 5.00529 17th 4.94549 18th 4.90205 19th 4.91258 20th It's a long term investment no doubt... as more discover that W7-64 bit disk that came for free, or can be downloaded more gain will be clocked. Still have to do it for my laptop, problem being, it's the only device that a few legacy 32 bit apps can run that refuse even when using the XP-32 compatibility mode... they load and exit without any notice. --//-- |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Hi SekeRob,
Think twice before losing your 32-bit compatibility. I have, and finally got so irritated that I recently boughtt a 4 GB dual-core that I intend to reformat as XP 32-bit once I rearrange my furniture. I have too many 32-bit legacy programs that I use and like which require 32-bit device drivers for internet connection. I could not get them to run on a VM under Windows 7 64-bit. Lawrence |
||
|
|
uplinger
Former World Community Grid Tech Joined: May 23, 2005 Post Count: 3952 Status: Offline Project Badges: 
|
Greetings all,
We have recently released a 64 bit version of Computing for Clean Water for Mac-Intel machines. This version will appear as version 6.42 on Mac-Intel 64 bit only. Thanks, -Uplinger |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
NICE! 64-bit Mac is ready, I have 8 cores coming atcha!
|
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
[Oops, wonder how I managed to enter 7.6 instead of 1.6 in the cache pref of me restarted quad], enough in the 64 BITS buffers to hit 1 year in the swiftest fashion. And the funny bit... WCG remembered that the machine was reliable, so a bunch of backend tasks of the "re-issue" type made all want to process high priority :O)
--//-- |
||
|
|
astroWX
Advanced Cruncher USA Joined: Sep 1, 2007 Post Count: 56 Status: Offline Project Badges: 
|
I don't understand the little I know about the 64- v. 32-bit issue but, for what it's worth, here is a quick-n-dirty comparison of my four Q9550 boxes (not overclocked; W7_x64 or Vista_x64; one with fastest time has 4Gig RAM, others have 8Gig; three have boinc 6.2.19, slowest has 6.10.58 and runs c4cw ver. 6.41).
CPU time: Ver 6.41 3:28 --> (runs only c4cw) Ver 6.40 3:21/3:23/3:23 --> (run a mix of c4cw and CPDN Beta) That seems to rain a bit on the "faster 64-bit" parade. As noted, "for what it's worth" and that probably isn't much. [5+ years to c4cw, so far] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
The techs [knreed] revealed that there is new server code coming [don't know when], that is going to test if a device performs better with 32 or 64 bits science apps i.e. there's awareness of that possibility. It does this by sending a lot of each and then compare and then decide if the 32 bit is what needs sending. Another grid has this in place and tests this by sending 80 of each. Don't know at all how this works for ''known'' member devices. Of course this comparing only works for devices that can do both. There are these days devices around that can only do 64 bit.
--//-- |
||
|
|
Ingleside
Veteran Cruncher Norway Joined: Nov 19, 2005 Post Count: 974 Status: Offline Project Badges:
|
The techs [knreed] revealed that there is new server code coming [don't know when], that is going to test if a device performs better with 32 or 64 bits science apps i.e. there's awareness of that possibility. It does this by sending a lot of each and then compare and then decide if the 32 bit is what needs sending. Another grid has this in place and tests this by sending 80 of each. Don't know at all how this works for ''known'' member devices. Of course this comparing only works for devices that can do both. There are these days devices around that can only do 64 bit. 80 seems to be on the large side... Atleast then it comes to adjusting the expected run-times, this happens after 10 validated results, so if not mistaken this will also be the case for starting to use the fastest application... If project uses up-to-date web-code, the info about the various applications, and which application is fastest in case of various bit-ness or GPU/CPU-ness will show-up for each device, so users can check this themselves. Example from SETI@home shows this info: http://setiathome.berkeley.edu/host_app_versions.php?hostid=5474637 SETI_Enhanced CPU : Average processing rate 16.692373562174 GPU: 797.6149750928 Astropulse CPU: 37.764120113081 GPU: 401.66996296419 Also, the server-code clearly uses the fastest resource for each application, since a test with re-issuing tasks, you'll only get 20 per asking from SETI, if you asks for both GPU and CPU you'll only get GPU-ones until you don't need any more GPU's.  "I make so many mistakes. But then just think of all the mistakes I don't make, although I might." |
||
|
|
astroWX
Advanced Cruncher USA Joined: Sep 1, 2007 Post Count: 56 Status: Offline Project Badges:
|
Thanks, SekeRob and Ingleside. Pertinent updates.
|
||
|
|
|