2020 Formula Boinc Sprints

Page 16 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Orange Kid

Elite Member
Oct 9, 1999
4,323
2,110
146

Fardringle

Diamond Member
Oct 23, 2000
9,184
753
126
A little over two thirds of all of the work downloaded since the sprint was announced has been returned and validated, and so far we're still maintaining a lead of about 160K points over TSBT. There are still almost 20K tasks in progress so that can definitely change, but hopefully you folks still have lots of pending tasks out there. I'm down to about 100 tasks waiting to be validated...

Edit: Actually, it looks like there were about 19K tasks out before the sprint was announced, so it's possible that most of the work downloaded for the sprint has been returned at this point, and there are a bunch of "regular" users still working on what they were already doing before the sprint. I guess we'll find out soon enough. ;)
 
Last edited:

Fardringle

Diamond Member
Oct 23, 2000
9,184
753
126
There are only about 2200 tasks still pending, and XtremeSystems had a nice dump over night, moving them about 50K ahead of TSBT.

But (and it's a really big but)....

Someone on our team also had a ridiculous dump over night. Free-DC hasn't updated yet, but looking at the stats on the project site, I believe a big chunk of it came from xii5ku (edit: and StefanR5R). Since there won't be any more work created before the end of the sprint, this will probably be the final rankings (with slightly different point totals):



Pos. Team Points Total credits
1​
TeAm AnandTech​
25
1,297,033
2​
XtremeSystems​
18
394,164
3​
The Scottish Boinc Team​
15
352,804
4​
UK BOINC Team​
12
336,718
5​
L'Alliance Francophone​
10
56,230
6​
Russia Team​
8
52,657
7​
Rechenkraft.net​
6
40,885
8​
BOINC@Poland​
4
15,424
9​
Planet 3DNow!​
2
12,709
10​
[H]ard|OCP​
1
11,136
 
Last edited:

StefanR5R

Elite Member
Dec 10, 2016
5,459
7,718
136
I was fortunate to receive work for almost 2 days for all of my computers (except some 4-core computers for GPU computing, which are currently powered down). Though, relative to their computing power, some computers got more work, some less. Yesterday evening I moved some of the work from still busy computers to already idle computers, and so they finished almost all at the same time today in the morning (i.e. last night in your time zone).

So far, it's evident that TeAm AnandTech was lucky to have a German branch. ;-)
[edit: ...considering the timing of work availability, shown in post #372.]
Most members of LAF, RKN, and P3D are obviously living in the same time zone as myself though, and yet they didn't use their chance to receive work as well as I did.

Another favorable circumstance was that the RakeSearch scheduler may send replicated tasks of one and the same workunit to hosts which are associated with the same user. If it hadn't done so, I would have received fewer tasks — but likely still enough for TeAm AnandTech to win this one (if this sprint goes through like it did so far).
 
Last edited:

Orange Kid

Elite Member
Oct 9, 1999
4,323
2,110
146
RakeSearch day two

1TeAm AnandTech251,302,868
2XtremeSystems18398,881
3The Scottish Boinc Team15355,192
4UK BOINC Team12341,355
5L'Alliance Francophone1056,953
6Russia Team853,559
7Rechenkraft.net641,392
8BOINC@Poland415,496
9[H]ard|OCP213,686
10Planet 3DNow!113,549

From FreeDC

biodoc9,82490,322100,146
crashtech17,67477,09394,767
xii5ku43,99415,85059,844
emoga9,44845,21154,659
Fardringle10,57036,62247,192
10esseeTony3,61621,64125,257
Orange Kid2,46616,14318,609
Icecold-Team Anandtech454,2054,250
zzuupp0373373
 

Fardringle

Diamond Member
Oct 23, 2000
9,184
753
126
Ah. Okey dokey.

Just trying to figure out why the daily totals aren't even close to the sprint stat totals so I was wondering if maybe someone was missing from the list..
 
Last edited:

Orange Kid

Elite Member
Oct 9, 1999
4,323
2,110
146
RakeSearch Final

1TeAm AnandTech251,306,564
2XtremeSystems18403,802
3The Scottish Boinc Team15357,944
4UK BOINC Team12344,785
5L'Alliance Francophone1058,112
6Russia Team857,304
7Rechenkraft.net643,417
8Planet 3DNow!416,711
9[H]ard|OCP216,563
10BOINC@Poland115,850

From FreeDC

xii5ku894,85643,99415,850954,700
crashtech12,46617,67477,093107,233
biodoc09,82490,322100,146
emoga669,44845,21154,725
Fardringle4,24910,57036,62251,441
10esseeTony7,4003,61621,64132,657
Orange Kid02,46616,14318,609
Icecold-Team Anandtech0454,2054,250
zzuupp00373373


Results from Belgium

  • Lewis HAMILTON1:24:08.761
  • Valtteri BOTTAS+8.448
  • Max VERSTAPPEN+15.455
 

Fardringle

Diamond Member
Oct 23, 2000
9,184
753
126
Interesting. They announced it two days early instead of just 1 day early.

Oh well, that just means 48 hours of bunkering instead of 24. :D

Edit: I see that it's also only a 48 hour sprint instead of the usual 72 hours...
 
Last edited:

biodoc

Diamond Member
Dec 29, 2005
6,257
2,238
136
Interesting. They announced it two days early instead of just 1 day early.

Oh well, that just means 48 hours of bunkering instead of 24. :D

Edit: I see that it's also only a 48 hour sprint instead of the usual 72 hours...

I had the same thought but @StefanR5R cleared it up for me.

"No, it's the usual 3 days:
starts Friday, 00:00 UTC = Thursday, 20:00 EDT
ends Sunday, 23:59 UTC = Sunday, 19:59 EDT"
 

Fardringle

Diamond Member
Oct 23, 2000
9,184
753
126
Oh. I see. He changed the way the start time is listed. The other recent sprints started and ended at 10 PM UTC. This one starts at Midnight. I didn't notice the time change so it messed up the time calculations in my head. :)
 
Last edited:

Fardringle

Diamond Member
Oct 23, 2000
9,184
753
126
I guess bunkering didn't work out so well for me, either. Thinking it was starting in two days, I told my computers to grab enough work to last more than that based on the estimated completion times. Turns out the estimates were wrong by a factor of 10, so I'm already out of work to do.

I guess it's time to follow that tutorial someone posted a while back to make another BOINC instance on the fast machines so they can try to grab some more work...
 

Fardringle

Diamond Member
Oct 23, 2000
9,184
753
126
I'm not planning to run multiples at the same time. Just wanted to have a second instance ready to grab more work for when my two faster machines completely run out of work (which they will do in an hour or two). Unless you think even doing that would cause problems...?

If so, I guess I could either just let those computers be idle for half a day, - or work on something else - or maybe set up a Linux VM to keep them busy.
 

emoga

Member
May 13, 2018
187
301
136
I'm not sure to be honest.

If I were in that situation, I would use a virtual machine instead of an another instance. (Preferably on a different operating system for good measure.)
 

Fardringle

Diamond Member
Oct 23, 2000
9,184
753
126
I've already played with a couple of Linux VMs on my 3900X computer, so I'll probably just do that. Thanks. :)
 

StefanR5R

Elite Member
Dec 10, 2016
5,459
7,718
136
I'm not planning to run multiples at the same time. Just wanted to have a second instance ready to grab more work for when my two faster machines completely run out of work (which they will do in an hour or two). Unless you think even doing that would cause problems...?
It's the same thing.

However, the finding from 2018 may no longer be true (Yoyo, the admin, sometimes reconfigures things), or paint an incomplete picture (there may be circumstances under which clients on the same host are still treated by the server as distinct). Speaking purely hypothetical for now.
 

StefanR5R

Elite Member
Dec 10, 2016
5,459
7,718
136
However, the finding from 2018 may no longer be true (Yoyo, the admin, sometimes reconfigures things),
It is still true; I tested it just now. (I tested in a way which did not cost me any completed work.)

or paint an incomplete picture (there may be circumstances under which clients on the same host are still treated by the server as distinct).
I tried one specific means to make the clients different from each other, and it did not prevent them from being collapsed into one by the yoyo@home server.
 

Fardringle

Diamond Member
Oct 23, 2000
9,184
753
126
I decided not to risk it. I just have the 3900X working on a different project until it's time to start reporting work for the sprint.
 

StefanR5R

Elite Member
Dec 10, 2016
5,459
7,718
136
Sprint #15, Yoyo@home, the first 19 hours:

1st 19 hours.png

(league 1, top 10 teams)

XS speculated successfully on the sprint project. The graph shows the first two of four hours of their initial dump. Since then they are running normally with 150...300 k/h, which is quite a bit better than every other team. That is, their speculative bunkering is not instrumental to their probable victory, but it gives them the opportunity to quit this sprint early and turn to whatever interests them more.