2020 Formula Boinc Sprints

[Ken g6]

By the way, if you were wondering what CERN might have done with your ATLAS WUs in the recent past...

ATLAS result addresses long-standing tension in the Standard Model

Researchers from the ATLAS collaboration explain their new measurement of "lepton flavour universality” – a unique property of the Standard Model of particle physics. (Video: CERN) The best-known particle in the lepton family is the electron, a key building block of matter and central to our...

home.cern

 

[Icecold]

My linux boxes are failing at the checkpoint, 10 minutes into crunching (Theory VM). Probably do not have virtualbox fully installed or something. Ah, well, good run from those who making all the points!

I have the exact same thing happen on one of my machines. 2 Different machines, same exact version of Linux Mint, installed Virtualbox the same way on both, etc. and one works fine and one fails every task 10 minutes in. Unfortunately it's the slower of the 2 machines that works fine. Definitely not a straight forward project to run.

 

[StefanR5R]

The final 18 hours:

 

[Orange Kid]

LHC@Home final

1	XtremeSystems	25	3,288,913
2	TeAm AnandTech	18	1,779,905
3	Gridcoin	15	1,568,943
4	The Scottish Boinc Team	12	1,015,322
5	Rechenkraft.net	10	838,502
6	Planet 3DNow!	8	418,897
7	UK BOINC Team	6	327,470
8	Overclock.net	4	192,408
9	L'Alliance Francophone	2	144,833
10	Czech National Team	1	86,297

From FreeDC

xii5ku	286,805	243,971	197,628	63,086		791,490
biodoc	105,861	117,121	128,716	47,780		399,478
wes.	191,098	145,457	26,869	24,618		388,042
crashtech	29,714	19,469	0	2,781		51,964
Orange Kid	21,247	12,102	0	1,729		35,078
Icecold-Team Anandtech	4,541	23,409	0	3,344		31,294
Fardringle	0	5,258	0	751		6,009
Peter	1,084	1,451	0	493		3,028
Ken_g6	2,772	0	0	0		2,772
far	0	366	0	180		546
Kiska	301	0	0	0		301
10esseeTony	0	108	18	18		144

and the British Grand Prix

• Lewis HAMILTON1:28:01.283
• Max VERSTAPPEN+5.856
• Charles LECLERC+18.474

 

[biodoc]

Second place works for me. Nice job all! Thanks for the stats @Orange Kid and @StefanR5R .

 

[Ken g6]

Second place works for me. Nice job all! Thanks for the stats @Orange Kid and @StefanR5R .

Especially since we passed L'Alliance Francophone in the overall 2020 results!

 

[Fardringle]

Nice job, team! Makes me feel (slightly) less annoyed that I was only able to get a few low credit tasks..

 

[StefanR5R]

There was yet another British Grand Prix (the 70th Anniversary GP actually), at the Silverstone circuit on August 7–9:

• Max Verstappen .... 1:19:41.993
• Lewis Hamilton ............ +11.326
• Valtteri Bottas ............... +19.231

And they took sanitarian risks by going to Spain on the Circuit de Barcelona–Catalunya on August 14–16:

• Lewis Hamilton ...... 1:31:45.279
• Max Verstappen ........... +24.177
• Valtteri Bottas ............... +44.752

 

[Fardringle]

Put on your fuzzy leg warmers and headbands and start stretching to get ready for the next sprint in just over 48 hours.

Spandex shorts are optional, and for some of us probably prohibited.

 

[Fardringle]

The next sprint has been announced. It is Rakesearch. The project doesn't have a lot of work available, so it would be best to stock up soon before it gets cleaned out.

 

[Fardringle]

Already completely out of work. I managed to grab some, but not enough to keep all of my computers busy. Hopefully the project will add more tasks soon.

 

[Markfw]

Sorry to be a killjoy, but why do they have sprints, where you can't get any work ? Why not pick projects more than once a year that DO have work ? Like Rosetta, and WCG for example.

 

[Fardringle]

Agreed. Rakesearch has frequently been completely out of work in the past, with only a relatively small number of active participants. Not a project that can support a mass sprint.

I would at least expect the folks running the Formula BOINC event to let project managers know ahead of time so they can expect the increase in traffic and can plan ahead with more work. Or say that they don't want to (or can't) be the sprint project.

 

[StefanR5R]

The sprint project announcement was during the night, local time of the project admin. He added new work now in the morning.

Sorry to be a killjoy, but why do they have sprints, where you can't get any work ? Why not pick projects more than once a year that DO have work ? Like Rosetta, and WCG for example.

One of the goals (or the goal) of Formula BOINC is to bring donors' attention to small projects. From this perspective, it would not make sense to get stuck into a cycle of WCG/ Rosetta/ Einstein/ PrimeGrid/ WCG/ Rosetta/ Einstein/ PrimeGrid/ WCG/ Rosetta/ Einstein/ PrimeGrid/...

I would at least expect the folks running the Formula BOINC event to let project managers know ahead of time so they can expect the increase in traffic and can plan ahead with more work. Or say that they don't want to (or can't) be the sprint project.

In the past, Sébastien was repeatedly asked to do this. I don't know whether or not he still doesn't communicate with projects. The 6½ hours of lack of work may have happened despite communication, if there was any.

 

[Fardringle]

It won't last long, but yes, there is more work available at Rakesearch right now. Get it while you can!

 

[StefanR5R]

Some random observations:

When the previous sprint at LHC@home was announced, people had speculative bunkers with 43 thousand RakeSearch results ready. Not so many this time.

RakeSearch is configured such that a user can be its own wingman.

Current server_status says that the "SAT-CMS-based search for orthogonal pairs of DLS of order 10" project is 91 % complete by now. Let's see how this advances during the sprint.

Last year, and the year before that too, we made 9 M in each of the RakeSearch sprints. If we repeat this this time, we might make it into the top ten of this year's RakeSearch marathon. Right now we are 15th. It may be close, because there are several teams in front of us in the marathon who usually take part in sprints too.

 

[StefanR5R]

Current server_status says that the "SAT-CMS-based search for orthogonal pairs of DLS of order 10" project is 91 % complete by now. Let's see how this advances during the sprint.

Uh oh...
One of the project developers(?)/admins(?) practically stated that those last 9 % of work were handed out last night/ this morning.

We were not aware of the sprint, sorry about that. Unfortunately, all workunits of the current experiment have already been generated, so we have nothing to generate at the moment. It would be better to choose another project for this purpose.

I wonder if Sébastien will let this sprint run its extraordinarily short course, or cancel it retroactively like the sprint at the bogus Wanless project... :-|

I would at least expect the folks running the Formula BOINC event to let project managers know ahead of time so they can expect the increase in traffic and can plan ahead with more work. Or say that they don't want to (or can't) be the sprint project.

In the past, Sébastien was repeatedly asked to do this. I don't know whether or not he still doesn't communicate with projects.

Well, at least now we know.

 

[Fardringle]

I was able to bunker about 50K worth of tasks (not enough to last until the sprint starts, but not too bad) and have about 5 or 6K worth sitting in "validation pending" from before the sprint was announced. I am sure some big producers snagged a lot since the available work disappeared really fast. I guess we'll just have to wait and see how much of that was done by members of our team..

 

[Fardringle]

Good start after the first hour update. I just dumped everything that I had, and most of it should show up on the next hour update, with several hundred more tasks pending validation.

It'll be interesting to see what other teams were able to snatch before the available work was gone, since there won't be any more...

1​	TeAm AnandTech​	25​	140,590​
2​	UK BOINC Team​	18​	73,722​
3​	The Scottish Boinc Team​	15​	57,670​
4​	XtremeSystems​	12​	34,790​
5​	Russia Team​	10​	12,588​

 

[Fardringle]

Still looking really good.

1​	TeAm AnandTech​	25​	260,482​
2​	UK BOINC Team​	18​	85,795​
3​	The Scottish Boinc Team​	15​	72,621​
4​	XtremeSystems​	12​	44,792​
5​	Russia Team​	10​	16,943​

 

[Fardringle]

Not much changing now so I'll stop spamming.

 

[StefanR5R]

Evidently, some just need more time to get rid of their stash:

 

[StefanR5R]

Apropos, I keep forgetting it: Is server_status.php showing tasks in progress like it says, or is the figure there actually workunits in progress?

 

[Fardringle]

12 hour update. Sorry I don't have individual user stats. Hopefully someone is tracking those...

Pos.	Team	Points	Total credits
1​	TeAm AnandTech​	25​	315,905​
2​	The Scottish Boinc Team​	18​	158,383​
3​	UK BOINC Team​	15​	146,945​
4​	XtremeSystems​	12​	103,331​
5​	Russia Team​	10​	32,131​
6​	L'Alliance Francophone​	8​	28,225​
7​	Rechenkraft.net​	6​	12,860​
8​	BOINC@Poland​	4​	9,844​
9​	Planet 3DNow!​	2​	7,943​
10​	[H]ard|OCP​	1​	5,010​
11​	Czech National Team​	-​	4,095​
12​	BOINC@AUSTRALIA​	-​	2,734​
13​	Team China​	-​	2,138​
14​	Overclock.net​	-​	1,902​
15​	Sicituradastra.​	-​	1,751​
16​	Gridcoin​	-​	1,553​
17​	Crunching@EVGA​	-​	1,435​
18​	SETI.Germany​	-​	1,426​
19​	Team 2ch​	-​	870​
20​	SETI.USA​	-​	320​
21​	Boinc@Taiwan​	-​	101​

 

[Kiska]

Apropos, I keep forgetting it: Is server_status.php showing tasks in progress like it says, or is the figure there actually workunits in progress?

I believe its "Tasks in progress", lemme check that now

Yes I just check its a single task, so "Tasks in progress"