WCG Christmas Race 2012

M

metallicafan

[H]ard|DCer of the Month - May 2010
Joined
Mar 30, 2005
Messages
2,205
It almost December and to World Community Grid fans that means its once again time for the WCG Christmas Race! Every year in December the World Community Grid team from Rechenkraft.net hosts a Christmas Race for a fun holiday challenge between WCG teams. The challenge is run through the WCG grid website and through their own stat updates so its a fair and friendly race for the good of the grid around the holidays. This year's race is a Points Challenge meaning that a team's score will be the total points from the teams completed work units, just as you would expect.

The [H]orde has participated in the race since 2008 and has always had great support form the DC subforum. Here are link to the results of past years: 2008, 2009, 2010, and 2011.

Here are some links for further information:
WCG Homepage
Christmas Race 2012 Results Page (updated once per day I believe)
Current Projects and Research
BOINC software download
GPUs NOT supported (and FAQ)
FreeDC HardOCP Stats Page

There are several great projects currently running on the grid. Universities are researching cures for diseases, ways to better filter water, renewable energy and even protein folding.

This year there is a slight new wrinkle to the grid. Until a month or so ago the WCG only used CPU applications. Recently the Help Cure Cancer project released an OpenCL GPU application. I have been running this application for the past month or so and its has been very stable on my Nvidia card and I've been told it performs even better on AMD cards. Now it seems the project is probably out of GPU work until after Thanksgiving (so dont worry if you dont get any GPU work right away). So with both CPU and a GPU application running on the Grid the [H]orde should be all set for a great showing! Anyone who wants to join in the fun (even if its only temporarily) would be more than welcome!

BOINC Client software Install Instructions
1. Go to this webpage to Register for an account and Select your projects: http://www.worldcommunitygrid.org/reg/viewRegister.do
Dont forget to join the HardOCP team #1411 (be careful as there are some old inactive [H] teams with similar names)
2. You can download the software from the WCG page but I would recommend getting it from the BOINC Download page listed above.
3. Once you have downloaded the software, go ahead and install the software. I would select all the defaults in the installer except I would select Advanced options and unselect the "BOINC Screensaver" option. You guys know the drill here regarding screensavers and performance. This is the [H]ardforum after all.
4. Then it will probably ask you to reboot and then the software will launch upon the reboot. Go to the Advanced View of the software, click on Tools and Add a Project. Select the WCG project and then enter your username and password when asked.
5. Thats it. The client should then go ahead and runs checks and performance benchmarks, download the client apps and WUs and start crunching.

A big thanks in advance to everyone who participates! If you have any questions let me know. Im sure there with be several BOINC guys monitoring this post and can help with install and config questions. I'll try and update this top post with any additional links/info that may be helpful. Crunch on!

Edit: The race runs from Dec1st to Dec 24th.
 
Last edited:
Dec 1 - 1.75M points - 14th place - 275 days RunTime
Dec 2 - 1.95M points - 14th place - 343 days RunTime
Dec 3 - 2.78M points - 12th place - 441 days RunTime
Dec 4 - 2.55M points - 12th place - 441 days RunTime
Dec 5 - 2.77M points - 12th place - 430 days RunTime
Dec 6 - 2.99M points - 12th place - 443 days RunTime
Dec 7 - 2.86M points - 11th place - 448 days RunTime
Dec 8 - 3.22M points - 11th place - 452 days RunTime
Dec 9 - 3.43M points - 11th place - 465 days RunTime
Dec10- 3.22M points - 11th place - 474 days RunTime
Dec11- 2.48M points - 11th place - 370 days RunTime
Dec12- 2.32M points - 11th place - 341 days RunTime
Dec13- 2.22M points - 11th place - 320 days RunTime
Dec14- 2.53M points - 11th place - 349 days RunTime
Dec15- 2.63M points - 11th place - 356 days RunTime
Dec16- 2.60M points - 11th place - 353 days RunTime
Dec17- 2.76M points - 11th place - 389 days RunTime
Dec18- 2.89M points - 11th place - 413 days RunTime
Dec19- 2.53M points - 11th place - 373 days RunTime
Dec20- 2.57M points - 11th place - 372 days RunTime
Dec21- 2.55M points - 11th place - 354 days RunTime
Dec22- 2.65M points - 11th place - 361 days RunTime
Dec23- 2.45M points - 11th place - 355 days RunTime
 
Last edited:
It is not just recommended but highly recommended to download the 7.0.36 or 7.0.38 clients. There are a lot of known issues with earlier version 7 client. Some users have found better experiences with the newer clients and they seem to handle GPU scheduling better.

On a side note, WUProp is an NCI (non-cpu intensive) project that can be ran alongside WCG units. It basically gathers stats on how OS, CPU's, etc... run at each project with each app. [H] has a low commitment at the project currently but it doesn't take away your commitment to WCG.

OProject and FreeHAL have NCI apps too. This would be a great time to get everyone to delve into those as well.

32bit:
http://boinc.berkeley.edu/dl/boinc_7.0.36_windows_intelx86.exe 7.0.36
http://boinc.berkeley.edu/dl/boinc_7.0.38_windows_intelx86.exe 7.0.38

64bit:
http://boinc.berkeley.edu/dl/boinc_7.0.36_windows_x86_64.exe 7.0.36
http://boinc.berkeley.edu/dl/boinc_7.0.38_windows_x86_64.exe 7.0.38
 
well that sucks was hoping my 8800GT's would be able to run the gpu WU's.. guess no WCG for me this winter.
 
I guess that I should mention that the Race begins on Dec 1st and runs through Dec 24th. If you are new to WCG, its best to start a couple days before the 1st to build up your daily points.

Also there is once again GPU work for the Help Cure Cancer project.
And yes EVGA has a team (they average about double our production so it will take a little works to be competitive)! :D

Time for a roll call. I should have about 16 CPU threads and 288 CUDA cores. :)
 
24+ CPU cores and 3 GT 430's. I also have borged machines jumping in and out.
 
ive got 6 full-time and 2 part time CPU cores i have switched over to the [H] i might be able to bring 6 more CPU cores full-time online if im industrious enough
 
I will suggest that [H] don't rely 100% on GPU work due to previous concerns of work flow availability. If you use an app info file, make sure you use one that allows at least one other science. That way if GPU runs dry, you will still have CPU work coming in. I would also suggest a zero redundancy CPU app like GFAM as well. That way you don't have to wait on anyone else to finish their work to get credit.
 
Last edited:
I would suggest any new WCG crunchers start now and not wait until Dec 1st. It will take a handful of validated units for each project before zero redundancy kicks in for those devices. Most of the projects are zero redundancy with the exception of Human Proteome Folding - Phase 2, Help Conquer Cancer and Discovering Dengue Drugs - Together - Phase 2.

https://secure.worldcommunitygrid.org/help/viewSearch.do?searchString=redundancy
 
pstang said:
ive got 6 full-time and 2 part time CPU cores i have switched over to the [H] i might be able to bring 6 more CPU cores full-time online if im industrious enough
Click to expand...
Great to have you crunching with us pstang! Thanks for the help!
 
Well there you have it. Plenty of GPU work for the Christmas Race! I can already see the points on the upward trend! Thanks guys! I have one more box to bring online later this week.

Anyone else want to join us? Anyone who might not be folding with AMD GPUs - these will perform great on the WCG!
 
I have decided to help you guys out for your christmas race, but I have some questions. First thing I am trying to figure out is how to get the program to use 100 percent of my processor. I went to preferences and made sure on multiprocessor, use 100 percent of processor and 100 percent of cpu time. It is constantly switching between 100 percent cpu and 50 percent cpu load.

The next thing I am trying to figure out is how to get it to use my gpu also. I have Use GPU while computer is in use checked. I have made sure to select use all projects in my user control panel on the WCG website. It has yet use my gpu at all. I have an HD 5850 graphics card. I was under the impression that AMD graphic cards worked great in bonic.

Then my last question. I am also going to load this program on my HTPC once I have all the bugs figured out. I have a HD 4770 in there. I have been trying to find information on if that graphics card would be supported in bonic. In the wiki it said HD 5000 or higher but just trying to confirm.

Thanks for any help you can give.
 
1. Make sure under your preferences where it says... While processor usage is less then You change this from the default 25 to something much higher. Or if you want a 100% dedicated, put 0 here. That will disable the option. Also, you have the option to run BOINC at all times or just when the pc is idle. The check box for While in use will need to be checked if you want it to always run.

2. There is an additional check box in the WCG preferences that says If my computer can process work on my graphics card.... make sure that is checked.

3. You can check the list of known cards here: https://secure.worldcommunitygrid.org/help/viewTopic.do?shortName=GPU#610 I do not believe the 4770 works at WCG. It should work at some other projects though.

We are also starting a thread of How To's here: http://hardforum.com/showthread.php?t=1729016

Also, keep in mind that the WCG HCC GPU Wu's will use a full CPU core/thread for each GPU unit. It wont actually use it 100% but will prevent any other work from loading into the core/thread while the GPU is active. They did this to guarantee the work units always had access to the core/thread when it was ready to start or finish. You do not need to make any changes for this to work. WCG programmed the apps to do this intentionally.
 
Last edited:
I also recommend attaching to this project while using BOINC. It is a non-CPU intensive app and will barely touch the CPU at all. Our team has a small presence there currently.

http://wuprop.boinc-af.org/

Other projects with non-CPU intensive apps are:
http://www.freehal.net/freehal_at_home/
http://oproject.info/
However, these two projects will require you to log into your account at their page and select to run the NCI versions. FreeHAL has CPU intensive and non-CPU intensive and OProject's NCI app is the ALX application.
These two projects are NCI but require additional hardware:
http://radioactiveathome.org/boinc
http://qcn.stanford.edu/sensor/
 
dcds1...another thing to keep in mind that with AMD cards you will need more then just drivers for the OpenCL projects to work. On the How To's link I mentioned some of the different versions of Catalyst. The newest Catalyst driver packages come with the SDK in the install pack. Just make sure it is selected when you do the install.
 
Thanks for the help Gilthanis. The program started running 100 percent full time on its own. I am not sure why it was cycling for awhile there. Under the Tasks tab I have projects listed as Ready to start, so maybe one of those might be a gpu project. But I am just happy that it as least using my full processor. I plan on adding my htpc as soon as it finishes the F@H project that it is working on.

What is the purpose of the wuprop program. I see that it collects stats from the bonic program. Are they trying to collect to create a huge database?
 
Pretty much. You can search their results based on processor and GPU for each project and each app at that porject and can filter based on OS.

http://wuprop.boinc-af.org/results/delai.py

Since apps keep getting updated or changed, the stats need constantly updated as well.
 
Our team name at that project is slightly different then WCG. Make sure you add the proper one.
[H]ard|OCP
 
dcds1 said:
I have decided to help you guys out for your christmas race, but I have some questions. First thing I am trying to figure out is how to get the program to use 100 percent of my processor. I went to preferences and made sure on multiprocessor, use 100 percent of processor and 100 percent of cpu time. It is constantly switching between 100 percent cpu and 50 percent cpu load.

The next thing I am trying to figure out is how to get it to use my gpu also. I have Use GPU while computer is in use checked. I have made sure to select use all projects in my user control panel on the WCG website. It has yet use my gpu at all. I have an HD 5850 graphics card. I was under the impression that AMD graphic cards worked great in bonic.

Then my last question. I am also going to load this program on my HTPC once I have all the bugs figured out. I have a HD 4770 in there. I have been trying to find information on if that graphics card would be supported in bonic. In the wiki it said HD 5000 or higher but just trying to confirm.

Thanks for any help you can give.
Click to expand...
Great to have you along for the ride dcds1! Thanks for helping the cause. Sounds like you are all up and running but if you have any further questions let us know. Crunch [H]ard!

Looks like the WCG team is doing well. Our 1.3M points for today is the best day we have had since last years Christmas Race. And the Race hasnt even started yet. Lets get to 2M PPD!

Team points for the past 365 days:
http://www.worldcommunitygrid.org/t...y.do?teamId=BP5XNJBR9N1&numRecordsPerPage=365
 
I am all up and running now. I finally figured out the gpu problem tonight. There was still one tab on the World Community Grid website that had to be checked. But now my gpu is folding away too. So hopefully in a day or two, my results will start being validated. At the time of writing this only 11 results have been validated. I have 5 pages of results waiting on pending validation.
 
metallicafan said:
<snip>Time for a roll call.</snip>
Click to expand...
Well, good news - the 4 cores I just restarted on WCG at stock 2.83 GHz this month are getting a new Hyper 212 and bigger case this weekend, so I'm hoping for at least 25%-30% more ppd from it after OC, and the new heat sink came in, too - so the good old Q6600 (marked DOA in my sig) should get up and running in the next week, as well!

So count me in for 8 cores, and perhaps a few GPUs if I can get them working - as I'm upgrading from an old broken XP install to 7 (or maybe 8 on the HTPC to try it out)...
 
Well...I have switched a few of the borged systems over to do just WCG and NCI projects and have a couple part time crunching borged GPU's. 660 TI and a GT 640, but these are off and on.

I did get a phone call from a cousin to build another Hex core AMD with a 660 TI, but she doesn't have the money until tax return time. So, no good for the Christmas race.
 
It seems that my informal tests indicate that the GFAM (Go Fight Against Malaria) project might be scoring a little better than other CPU projects. Anyone else had similar results?

Having said that I would encourage everyone to still crunch towards the project they were most interested in. :)
 
I am primarily running SN2S and GFAM on my CPU's right now and am finding higher points on SN2S then GFAM. However, both of those projects were putting out better numbers then HPF2 on the same device. Either way, I'm running those two projects for a run time goal rather then points. If DDDT2 kicks in soon, I will be focusing full scale on that to try and get my fifth year goal in before it finishes.
 
Opening day and we are placed at 14th. I know we can do better. Crunching@EVGA is quadrupling our production on the first day.
 
I was just looking at our position myself. I think I would be happy if we could just double our current point output and make it into the top 10. But no matter how we end up doing, atleast it is all for good causes. And I know the F@H team is trying all they can do to keep their 1st place spot.
 
I should be adding 3 more cores from a work PC on Monday...but not sure if it will be able to run 24/7 yet.
 
I can add to the push for at least a week.
 
Gilthanis said:
I am primarily running SN2S and GFAM on my CPU's right now and am finding higher points on SN2S then GFAM. However, both of those projects were putting out better numbers then HPF2 on the same device. Either way, I'm running those two projects for a run time goal rather then points. If DDDT2 kicks in soon, I will be focusing full scale on that to try and get my fifth year goal in before it finishes.
Click to expand...
Thanks for the info. I an just finishing up a badge in DSFL and then since I havent done any SN2S yet I'll start on that project for a while. :)

Gilthanis said:
Opening day and we are placed at 14th. I know we can do better. Crunching@EVGA is quadrupling our production on the first day.
Click to expand...
We may not be in the Top 10 but our first day of 1.7M points was our best day in the past 365 days by about 200K. Perhaps we can make it to 2M PPD or beyond?

KMac said:
I can add to the push for at least a week.
Click to expand...
Good to have you along sir! Its very encouraging to see so many people participating!
 
I would set the goals a bit higher then that. GPU work has pretty well doubled my production and there are quite a bit of members here who have some pretty impressive GPU hardware at their disposal. I think 2 million would be a very simple task for this team.
 
nice bump in points today... we moved up a couple of positions... i wish i could get my last 4 cpu cores online, but that probably wont happen for another week or so
 
Wow! Great day today guys! 2.7M might be our single biggest points day in WCG ever. :)

Its great to see so many people contributing (big and small)! Kmac I gotta ask - What kind of hardware does one use to score 120K in a single day? In a day where I see many impressive scores on the list and lot of people boosting their normal production for the challenge, that is a very impressive feat!

Well done all. Crunch on!
 
metallicafan said:
...Kmac I gotta ask - What kind of hardware does one use to score 120K in a single day?...
Click to expand...

120K?... Wait for the update.

I asked the Commandos for help passing EVGA at F@H a while ago and you guys stepped up remarkably. Time to return the favor. Currently running ~130 cores, 2x4P, 1 SR-2 and a few desktops.

I was laid off a while ago so I had to back off on the DC to save on electricity for a bit, but I can still return a favor. Crunch and/or Fold on!!!
 
I knew there was some hidden muscle. Thanks Kmac!
 
Incredible KMac. I just got mowed over in a few days at over 1M boinc points. Thanks for the help.
 
You must log in or register to reply here.
Back
Top