| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: Discovering Dengue Drugs - Together
Thread: DDDT Single Redundancy Change |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 87
|
|
| Author |
|
|
knreed
Former World Community Grid Tech Joined: Nov 8, 2004 Post Count: 4504 Status: Offline Project Badges: 
|
About 11.5% of the workunits are falling into the first category at this time. This means that 88.5% of the workunits are being sent to computers deemed sufficiently enough reliable to do the work on their own.
----------------------------------------That means average replication at this time is about 1.12. This will go up as random sampling on the back end increases this. However, it looks like we should have average replication below 1.25. For a reference, we were at around 2.2 before this change. That means that we should be able to get almost 75% more workunits completed (for Discovering Dengue Drugs) with the same computing power (provided by the members). [Edit 1 times, last edit by knreed at Jul 30, 2008 6:51:41 PM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
knreed,
Thanks... so the next obvious questions would be.. how does one ascertain how "reliable" their machine is?
Wouldn't it be good for those donating their equipment to know if they aren't returning useful information? |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
The criteria are:
-----------------------------------------Returning valid results and keep eye on your Result Status pages: You know best! -If wanting to candidate for short deadline/rush work, keep a small buffer, under say 1 or 2 days. If you see your daily credits take a plunge or a message in your client log showing "reached quota", the servers have determined that your machine is not worthy. Sometimes if a client persists in sending 1 invalid result per day, bottom quota, you get a mail from Big Brother right hand. I'm pretty darn sure I wrote the same thing or similar a few days ago as did several other CA's. Maybe the admins delete our posts to keep us busy 
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I realise that you or one of the Techs have probably posted this somewhere that I can't seem to find Sek - so I apologise in advance - however I can't quite work out from the reference to credit allocation in point 4 of Kevins explanation above of how single redundancy works, exactly how the granted credit is calculated.
Am I correct in thinking that its something like after the short run of the application that computes a known result by the Techs - they then work out some kind of average speed of computation and credit claim and then look at the efficiency/speed etc of each claiming computer and apply granted credit according to some kind of formula? |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
There is a little benchmark right at the beginning of a job, every job, that computes a known piece of science. That and the variation for each DDDT benchmark is used to build a claim rating for a device forming part of the computation of credit. Let me dig out if anything was written over what was said in point 4.
----------------------------------------I've not done new DDDT yet, so cant say from hands on, but one CA achieved a claim to grant variation between 0 and 4.3% The workings for the new FAAH will be the same. I just know from the 43 beta I got of last test, the difference was minimal.
WCG
----------------------------------------Please help to make the Forums an enjoyable experience for All! [Edit 1 times, last edit by Sekerob at Aug 1, 2008 8:38:40 AM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Single redundancy simply means that a particular WU is not repeated on multiple machines. Basically the scientists have trust that the results that are sent back by computers crunching a particular WU will be correct and that they do not need to send out exactly the same WU to another computer to see if it gets the same result.
Multiple redundancy is used to ensure that results are reliable; quorum is the number of results need in order to validate WU. As for credit, as there is a single result, you will be given what your computer claims. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Thanks to you both for your replies...
There is a slight variation in what you both say regarding credit - Sek says it may be based on prior claim ratings for a particular computer along with a mini benchmark at the begining of each job, and stares says you get what your computer claims as its a single result. From what I have seen from results completed by some here already, the claim can be different to the granted credit, although not greatly so - so I imagine that Sek may be more on the button? |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
Hi Ady again,
----------------------------------------As what i wrote earlier it's definitely hinging on that little pre-job integrated benchmark of a reference piece of DDDT science. Do that well, you get equal or more than the claim.... do it not so well, you get a little less. Over time it's expected to converge, so at least the data from the beta test supported the theory. Anyway, i ran the chart machine this morning and on 1st day the credit average practically jumped back to the value it had before the switch and per-machine it should refine as more results are being returned. ttyl
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Sounds reasonable to me Sek, thanks again.
As you can probably imagine at some places they are very keen to know and understand the full ins and outs of the calculation method etc, so they can be assured they are getting what they should - perhaps if one of the Techs has a spare moment, they could pop here and post an explanation in somewhat laymans terms to keep the clamouring hordes happy  |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
I'm sure the clamouring hordes will be on/in our tail pipe if credit is deviating materially from claim. Mine did not in the beta run, the only sample of the 43 I had still listed remaining on my Result Status page made me appy.
---------------------------------------- BETA_ dddt0602h0452_ 101058_ 0-- 628290 Valid 07/21/2008 17:50:57 07/22/2008 05:48:44 1.77 27.1 / 29.3 The methodology for FAAH in that beta was same.
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
|