WCG - 15th Birthday Challenge

[Terminator_1]

Dear members of AnandTech ,

For Nov 16th to Nov 22nd 2019 SETI.Germany is arranging a challenge at World Community Grid on the occasion of the grids 15th birthday:

We would like to invite AnandTech to participate in the challenge and hope that your team and many members of your team will be joining and participting this event.
!!Be aware, only your WCG-Team-Captain can join your team to the challenge!!
https://www.worldcommunitygrid.org/team/challenge/viewTeamChallenge.do?challengeId=9700

More information and statistics regarding the challenge can be found here.http://www.seti-germany.de/wcg/1_en_Welcome.html


Your participation is very much appreciated.

Best regards
Terminator
SETI.Germany

 

[Markfw]

Lets do it !

 

[Wiz]

So who is this "WCG-Team-Captain"?

 

[Thebobo]

At least I can heat my house at that time of year.

We have a team? Or someone needs to make one for us?

 

[StefanR5R]

We have a team? Or someone needs to make one for us?

Here is our team page: TeAm AnandTech at WCG

So who is this "WCG-Team-Captain"?

@petrusbroder is the captain.

Lets do it !

Indeed!

 

[StefanR5R]

Links to earlier WCG Birthday challenge threads: 2018, 2017, 2016

Important dates:

• Today --- Begin to prepare.
  Make sure you are member of the TeAm.
  Have all your computers return a few results results and get them validated. After that, they are trusted by the project servers and can receive many tasks into the work buffer.

• Wednesday, November 06, 00:00 UTC = Tuesday, November 05, 16:00 PST --- Start to bunker.
  Choose subprojects with 10 days deadline.
  Subprojects with 7 days deadline can be bunkered from Saturday, November 09, 00:00 UTC = Friday, November 8, 16:00 PST on.

• Saturday, November 16, 00:00 UTC = Friday, November 15, 16:00 PST --- Start of the Birthday challenge.
  (convert time and date into local time)

• Friday, November 22, 23:59 UTC = Friday, November 22, 15:59 PST --- End of the Birthday challenge.
  (convert time and date into local time)

 

[Markfw]

OK, in my PDT time zone we are 7 hours earlier than UTC, so I can start bunkering 7 day tasks at 5 pm Friday Friday Nov 8th ?

I could have either 232 or 248 cores and 464 or 496 threads on that day (depends on when the 3950x comes out and availability.).

This year I am bunkering all cores on wcg for 7 days. At my current rate, that could be a 2 million point bunker. I am only doing one. But my total I hope is then 4 million points for the week.

 

[StefanR5R]

OK, in my PDT time zone we are 7 hours earlier than UTC, so I can start bunkering 7 day tasks at 5 pm Friday Friday Nov 8th ?

Oregon will return from daylight time to standard time on the first Sunday in November (11/3), from then on you'll be at UTC-8. (00:00 UTC translates to 16:00 PST on the prior calendar day then.)

 

[Markfw]

I was browsing last years thread, and it seems that SCC gets way better ppd. I changed my preferences to that. If I am wrong, let me know now please.

 

[StefanR5R]

As per current projections, only Mapping Cancer Markers (MCM) and Microbiome Immunity Project (MIP) will have a continuous supply of work.

Recent status of the subprojects:

Microbiome Immunity Project (forum)

Here are notes from today's monthly call with the MIP researchers.

1. They are working on creating more batches--generally they like to have at least 2,000 available for download, but they'd like to have a larger backlog going forward. The first step (which they're working on) is pulling more targets to look at.
2. They're putting an informal paper up on a pre-print server (not an official publication) that gives an overview of how the researchers plan to use the MIP data. As soon as the paper is up, we'll publicize it in the MIP forum. This paper is considered a preface (or introduction) to the three formal papers that they've talked about in their project updates.
3. The three upcoming papers will focus more on protein function prediction and biology.
4. The researcher in Poland gave a high-level talk about the project a few weeks ago at a local meeting.

Update on work unit progress (as of October 1):
Available for Download: 1,730 batches*
In Progress: 9,925 batches (16,285,432 work units)
Completed: 22,1948 batches total, 7,749 in the last 30 days, 258/day average

*NOTE - The researchers are building more batches--this is just the number available for download as of the date of this post. The researchers do not have an estimate of how many total batches they will be submitting for the project.

(source)

Smash Childhood Cancer (forum)

Our monthly call with the researchers just ended.

1. The research team has tentatively chosen a new target to investigate using WCG. The new target could be promising for a number of germ cell tumors. We'll provide more information as we get it from the researchers. The next steps will be finalizing the target selection, building initial work units, alpha testing, beta testing, and then releasing work units to all volunteers who have signed up for the project.
2. The researchers are applying for a small grant to help cover the costs of working on the new target, and the WCG team will send a letter of support for their application. (We do this frequently for active research projects).
3. They're getting ready to do lab testing on some promising compounds--this is in the purchasing and permissions stage.

Work unit status as of October 9:

21 batches (78,1641 work units)
Completed: 3,665 batches - 302 in the last 30 days - average of 10 per day

(source)

OpenZika (forum)

We had our monthly call with the research team today.

1. The researchers confirmed that they've sent the last of the work units to us. Based on our tech team's estimated projections, there's about two weeks worth of work to be processed.
2. One of the researchers will be speaking at a conference in the US in early December.
3. They have several papers in the works--mostly in the writing stage.
4. They're making plans for analyzing their data--the project produced a huge amount of information.


Work unit status (as of October 14):

Available for Download: 0
In Progress: 6,807 batches (7,753,614 work units)
Completed: 457,567 batches - 6,155 batches in the last 30 days - average of 205 per day

(source)

Help Stop TB (forum)

We just wrapped up our monthly call with the HSTB research team.

1. The WCG tech team discussed other uses of Gromacs (the software the researchers are using for the project) with the scientists and got their opinions on how Gromacs might possibly be used for other volunteer computing projects. (We don't have any such projects in the works, we've just had questions from other scientists in the past.)
2. We also discussed their ongoing search for another team member, which is made difficult by the fact that applications for PhD and postdoc positions are down by ~75 percent. We hope there will soon be some good news!

Status of work units as of October 9:

In Progress: 62 batches (6,200 work units)
Completed: 22,503 batches - 139 in the last 30 days - an average of 4.6 per day

(source)

FightAIDS@Home - Phase 2 (forum)

Just wrapped up our monthly call with the FAAH2 research team.

1. They are almost completely done with the current work units and ready to begin analysis of the data they've received from WCG.
2. There is about 1-2 weeks of work (ESTIMATED) left for this current group of work units. Once they've run out, there will be a pause. At this point, we're not sure how long the pause will be.
3. Their analysis process will probably take a few weeks if everything runs smoothly. The purpose of the analysis is to see if they can refine the docking targets further.
4. The FAAH1 researchers may have further targets to send to the FAAH2 researchers that might be run on WCG in the future (they were not present during this month's call). The two groups are going to touch base over the next few weeks about this.

Work unit status as of October 9:

In Progress: 99,883 work units
Completed: 3,878,277 in the last 30 days - 12,9276 average per day

(source)

Mapping Cancer Markers (forum)

We had our monthly call with the MCM research team today.

1. They have a draft of the dataset for sarcoma work units, and hope to send us some test work units some time next week. Once we receive them, WCG's team will do alpha testing on these unit over the course of this month. We'll let everyone know when they go into beta testing, hopefully in November.
2. We let them know about about a new team created by a teenager in the US with an ambitious crunching goal.
3. We discussed which cancers the project might concentrate on in the future, after the sarcoma work is completed. The researchers aren't close to making any decisions yet--just very early brainstorming. They will probably choose a cancer related to the cancers they have already studied (lung, ovarian, and sarcoma).

Status of work units as of October 9:

Available for Download: 212 batches
In Progress: 812 batches (6,102,542 work units)
Completed: 54,052 batches- 758 in the last 30 days - average of 25.3 per day
Estimated backlog: approximately 8 days

(source)

 

[Markfw]

OK, I have selected all 3 now. Ready to switch over at the time. Still have 100k/day ppd on it.

 

[TennesseeTony]

OK, in my PDT time zone we are 7 hours earlier than UTC, so I can start bunkering 7 day tasks at 5 pm Friday Friday Nov 8th ?

I could have either 232 or 248 cores and 464 or 496 threads on that day (depends on when the 3950x comes out and availability.)

Mark, you will need to have multiple clients per machine to keep you supplied with work, due to the 1000 task limit.

 

[Markfw]

Mark, you will need to have multiple clients per machine to keep you supplied with work, due to the 1000 task limit.

I did some quick math. at 4 hours per task thats 6 tasks per thread per day times 16 tasks per box, that 96 per day. So 7 days is under the 1000 limit ? Thats 30 threads*6=240 or 4 days.

 

[StefanR5R]

Here is what else will be going on before, during, and after the WCG Birthday party:

         October | November                                                                       November | December
               Th Fr Sa Su Mo Tu We Th Fr Sa Su Mo Tu We Th Fr Sa Su Mo Tu We Th Fr Sa Su Mo Tu We Th Fr Sa Su Mo Tu
               31 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 01 02 03
 WCG Birthday                    __ __ __ __ __ __ __ __ __ __ ▄▄ ▄▄ ▄▄ ▄▄ ▄▄ ▄▄ ▄▄                                   WCG Birthday
Formula Boinc  __ ▄▄ ▄▄ ▄▄ ▄                             __ ▄▄ ▄▄ ▄▄ ▄                           _ _▄ ▄▄ ▄▄ ▄         Formula Boinc
    PrimeGrid      ▄ ▄▄ ▄▄ ▄▄ ▄▄ ▄▄ ▄▄ ▄▄ ▄▄ ▄▄ ▄                                                                     PrimeGrid
 Folding@Home                                                                                               ▄▄ ▄▄ ▄▄  Folding@Home

 

[petrusbroder]

The TeAm has joined the challenge.

 

[[H]Coleslaw]

At least they added their new sub project today as well.

 

[StefanR5R]

Thanks. This must have happened minutes after I last reviewed the "My Projects" settings. :-) Changed them now once more.

They upped the minimum system specs for this; I wonder what the RAM spec implies for a many core computer...

 

[biodoc]

I picked one up in my que on a 3700X. I'll try to monitor RAM usage when the task starts (estimated completion time is 11:04 (hrs/mins).


Application: Africa Rainfall Project 7.27
Name :ARP1_0002630_000
Received: Wed 30 Oct 2019 09:09:11 AM EDT
Report deadline: Wed 06 Nov 2019 08:09:10 AM EST
Estimated computation size: 206,976 GFLOPs
Executable: wcgrid_arp1_wrf_7.27_x86_64-pc-linux-gnu

 

[crashtech]

Looks like I can run the new project on some, but not all of my machines. So setting up a separate location may be necessary if it's a well-paying project.

 

[StefanR5R]

By default, the device profiles at the WCG web site are set up to allow only a single ARP1 task in progress:



Users can increase this if they know that their computers are capable of running more of these at once.
(source: WCG forum)

 

[biodoc]

The ARP task is using 713 MB of RAM.

 

[StefanR5R]

Until recently, before OpenZika concluded, those team who had lots of Linux and were not afraid to use it, had an advantage in WCG contests over teams whose users mostly clung on to Windows. I suppose this advantage is mostly gone now. :-|

 

[[H]Coleslaw]

I have work units buried in what is left of my que known as the land of OZ. Anyone know if Linux performs better on the new project yet?

 

[StefanR5R]

This is going to take a while to determine, for two reasons: 1. Long work units with seemingly variable and unpredictable run time. 2. WCG's credit estimation may be in flux for a while until it all settles in.

(I for one don't even know yet whether it's quorum=2, or adaptive validation.)

 

[StefanR5R]

• Wednesday, November 06, 00:00 UTC = Tuesday, November 05, 16:00 PST --- Start to bunker.
  Choose subprojects with 10 days deadline.
  Subprojects with 7 days deadline can be bunkered from Saturday, November 09, 00:00 UTC = Friday, November 8, 16:00 PST on.

• Saturday, November 16, 00:00 UTC = Friday, November 15, 16:00 PST --- Start of the Birthday challenge.
  (convert time and date into local time)

• Friday, November 22, 23:59 UTC = Friday, November 22, 15:59 PST --- End of the Birthday challenge.
  (convert time and date into local time)

I woke up in the middle of the night, for no particular reason.