2019 Formula Boinc Sprints

[ao_ika_red]

@ao_ika_red, good to see you being around!

Thanks for the shout out, Stefan. I wanted to do it under the radar, but O.K.'s stats wouldn't allow me to do that. It's good to be back.

 

[VietOZ]

nice drop @StefanR5R

 

[VietOZ]

Congrats on 3rd with a thin resources, guys! Well played at the end!
side note: @StefanR5R you're my wing man (and SaM's too). XS did 7.2M in that last hour drop just thru Mr.Pending lol
My pending went down 2k and SaM went down 2k+

 

[Orange Kid]

Universe@home final

WOW some massive dumps went since last I looked. Had to check three times to be sure I was looking at the right sprint stats.
Third place, mighty impressive, to say the least. WELL DONE ALL.




From FreeDC (obviously very incomplete picture)



Japan F1 results

• Valtteri Bottas 1:21:46.755
• Sebastian Vettel +13.343
• Lewis Hamilton +13.858

 

[ao_ika_red]

From FreeDC (obviously very incomplete picture)

Yeah, at least one protagonist is missing. Using GDPR for tactical gain perhaps?

 

[StefanR5R]

Using GDPR for tactical gain perhaps?

Nope, just being tardy with the reporting of results.

--------
Thanks to Sébastien for fixing the early stats issue quickly, and also thanks to the U@h admin (Krzysztof probably) for putting up a new certificate quickly as well.

 

[Ken g6]

This sprint may have its ups and downs. It's Yoyo@Home!

 

[ao_ika_red]

Great way to start (end) the day.

 

[Orange Kid]

These are the top ten spots in the FB marathon.
We currently hold the 5th spot. With the sprints that are left, mathematically we have no ability to move up any places with sprint points only. Conversely we can be overtaken and loose places.
This being the case, we simply need to stay ahead of the team below us in the remaining sprints.
Hope this helps with everyone's allocation of resources for the remainder of the year.

1 ​	Gridcoin​	858​
2 ​	XtremeSystems​	661​
3 ​	SETI.USA​	576​
4 ​	[H]ard|OCP​	418​
5 ​	TeAm AnandTech​	276​
6 ​	SETI.Germany​	258​
7 ​	L'Alliance Francophone​	251​
8 ​	Czech National Team​	244​
9 ​	Rechenkraft.net​	168​
10 ​	Planet 3DNow!​	151​

 

[Orange Kid]

YoYo@ HoMe day one+
Third place !!! Nice.
Hard to find a task that will run well on machines seems to be the big down fall this sprint.



From FreeDC

 

[StefanR5R]

Apropos nostalgia. I am missing the Xansons Evenings.

A tiny bit.

 

[crashtech]

@StefanR5R , it was a pain in the butt, but a little bit exciting nonetheless to see who could spam the server the best!

 

[Orange Kid]

Yoyo@home day 2+
In third!!!! What a wonderful thing.
Will there be any bunkers? Hard enough to run this project let alone hid some on the side. We shall see. lets hope our spot stays firm.



From FreeDC

 

[VietOZ]

Apropos nostalgia. I am missing the Xansons Evenings.

A tiny bit.

@StefanR5R , it was a pain in the butt, but a little bit exciting nonetheless to see who could spam the server the best!

I don't really miss it. Even though, one of our teammate had created a script that can launch however many instances we want, auto attach to the project, get work, crunch, upload.
The script would check each instance, if it runs out of work then it'll move to the next instance. If upload got stuck (project backoff) then it'll auto retry. And it'll loop back and forth until there's no work left. All we needed to do was adjust a variable for how many instances we wanted to run, then run the script. Done ... went play games for a few hours and come back to check on it
One dude went as far as 500 plus instances on a single GTX 1080ti. That's 500 plus "fake" machines pounding the server asking for work lol

 

[StefanR5R]

This was obvious from the timing of XS's returning the bulk of results. :-)
The top man of SETI.Germany used to do something similar at Xansons, though not as extensive.

PS,
that nostalgic thought came to me yesterday on my way home from grocery shopping. Usually, Xansons began to send tasks when I was out to the supermarket. I had a few canary clients which checked for work availability early; as soon as they got their first task, all the other clients would be woken up and begin fetching work too, all on their own.

 

[Ken g6]

Well, I'm glad this particular sprint is over. I put one P2 instance on my TV recording machine. It turns out I've allocated probably too much RAM to each recording instance, so this project prevented me from recording more than one show at a time! Too bad I didn't figure that out until just now.

 

[StefanR5R]

Congratulations to @phoenicis who found 2 ECM factors during this challenge:
https://www.rechenkraft.net/yoyo/y_factors_ecm.php

 

[biodoc]

Cool! Congrats @phoenicis !

 

[Howdy]

Well, I'm glad this particular sprint is over.

You're not the only one........towards the ending hours my machine got flooded with the P2s. It was having error after error due to memory issues. They would run for 3 mins or so and then error putting it in the time out corner. I had to babysit it and then just gave up and swapped it to PG.

 

[Orange Kid]

Yoyo@home Final, finally
Third place!!! A fine finish to a really rough project. Well done all.



From FreeDC



Mexico F1 results
Hamilton came close to securing the title but will have to try again next week in the U.S.

• Lewis Hamilton 1:36:48.904
• Sebastian Vettel +1.766
• Valtteri Bottas +3.553

 

[StefanR5R]

The next sprint project announcement is due in less than 3 days from now (in 2d16h).

...And soon after that sprint starts, yet another PG challenge will be on, which in turn will cut into WCG B-day preparation. --- It's good though to have choices. :-)

 

[Orange Kid]

Cosmology@home is the next sprint.

 

[Ken g6]

Well, I got a few Cosmo VBox tasks, but the server seems to be out of tasks now, so I guess I'll wait and see if more show up tomorrow.

 

[StefanR5R]

This sprint may move us up a little in the Cosmology@home marathon. We are 9th there currently.

 

[StefanR5R]

I was just idly browsing through the discussion of the previous Cosmo sprint. IIRC, the server had low database performance at that time, and a slow work generator. Also, the camb_legacy validator didn't work for periods of days or weeks as it was segfaulting all the time. Marius got the legacy validator going again during the sprint, which effectively caused a huge involuntary bunker release from Gridcoin.

The database performance problem was addressed last December by a server upgrade. I don't know if the work generator is faster now. At least yesterday its output rate was no match for the prospective sprint contestants.