| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: Human Proteome Folding - Phase 2
Thread: Interesting eyeball analysis of results |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 7
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I did q quick look at @90 valid work units. Except for two units, I got about 2/3 of the credit requested for the remainder.
Something strikes me as statistically amiss. |
||
|
|
smeyer55
Senior Cruncher Joined: Feb 15, 2009 Post Count: 303 Status: Offline Project Badges: 
|
If you are running on an i7 series processor, they are known to overclaim the credits.
|
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
If you are running on an i7 series processor, they are known to overclaim the credits. The bulk of my machines are Core 2 Duos running OS 10.6.3. |
||
|
|
rilian
Veteran Cruncher Ukraine - we rule! Joined: Jun 17, 2007 Post Count: 1452 Status: Offline Project Badges: 
|
64-bit systems are known to overclaim credits as well ...
---------------------------------------- |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
I did q quick look at @90 valid work units. Except for two units, I got about 2/3 of the credit requested for the remainder. Something strikes me as statistically amiss. Yes, those 2 units are the outliers of your 90 odd. As noted in replies above: 1, I7 when running with Hyperthreading enabled take much longer on a task, so the real work per hour is not reflective of the 64 bit client benchmark. Try running without. 2. 64 bit benchmark has a rather overweight portion from the Integer part of the test (Dhrystone), whilst sciences such as HPF2 have a very high floating point content (Whetstone). If you investigate you'll notice that the client packages for 32 bit and 64 bit Whetstone portion of the benchmark give the near identical value. So, beyond the eyeball nothing is wrong, but the benchmark.
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Regardless of the problematic-but-not-a-problem benchmark(s), I get significantly less credit per hour of CPU time than from my other projects.
In order to maximize the efficiency of my meager contribution to WCG, it seems the simple-yet-elegant solution is to drop this particular project and re-allocate my not-Winderz machines to more harmonious and receptive WCG projects. _____________________________________________________________ "Thus it is written, thus it shall be." (Yul Brynner as the Pharoah in the Ten Commandments) |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
Once one measures contribution by credit per hour versus other WCG projects, then indeed go for the highest ROC. The benchmark/credit system is getting a complete overhaul, but if that is going to land in 2010 or ever, I really don't know. It's a steep task. Whether the revision will translate to getting equivalent per hour for HPF2 to e.g. FAAH is anyone's guess.
----------------------------------------
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
|