2020 Formula Boinc Sprints

[Fardringle]

And it looks like the current stats are going to be the ONLY stats for this sprint..

Hi - FB Sprint

Edit: Direct link to the post doesn't work apparently, but this is what bearnol posted in response to requests to turn the server processes back on and/or temporarily extend the deadline on tasks so the 150K or so tasks currently waiting to be returned won't simply be discarded with no credit:

Thank you all (again), for both your messages of support, and criticism.
I do recognise that there has been some frustration recently. I think this is mainly due to the increase in relative speeds of everyone else’s computers, for which I guess we should all be grateful. Hopefully we are all on the side of BOINC tho’ , in general. (And actually I am personally very grateful to Formula Boinc for bringing in some excellent crunching over the years).
With regard to the person who asked re cancelling current workunits: I’m not really planning to restart the server’s processing for a little while, to give the rate a chance to cool somewhat.
With regard to those of you seeking deadline extensions etc.: I’ve always believed that my project serves a role to hardware beta testers in providing a quick and easy validatory benchtest, so I actually prefer a quick turnaround ( tho I could be imagining that The same argument applies, with varying reasoning, to still having fairly short workunits, which I, and a number of others since the start of the project, have said they liked. Conceivably I made a small mistake releasing ARM apps which run somewhat slower , reducing the scope for increasing wu length, but I think on balance this added to the project (and the ARM chips continue to improve in speed )
If you all become enough of a pain protesting ( ) , or if my trusty little server does eventually bite the dust, then hopefully I will upgrade at some point [though I haven’t really done any live speed comparisons with my replacement candidate servers, so I don’t know exactly how much faster they would be], but in the meantime please recognise that 1) I am obviously always very grateful to any and all of you who have given me constructive crunching over the years, but that 2) I am just one person and operating on a limited budget.
On a constructive note: I refer you (once again) to the thread I linked to above, in case anybody feels like branching out in pursuit of these Mersenneplustwo factors!

 

[TennesseeTony]

So with sixty (60) points [2 tasks], we gain 8 points? Kewl.

 

[Fardringle]

So with sixty (60) points [2 tasks], we gain 8 points? Kewl.

Looks that way.

Sucks for everyone who jumped in and processed over 150K tasks in 12 hours or so before the sprint timer started, and none of those tasks will be credited..

 

[Orange Kid]

Wanless Mersenne +2 day two
Second day same as the first, another day older and still FUBAR.

1	UK BOINC Team	25	664
2	XtremeSystems	18	620
3	L'Alliance Francophone	15	397
4	OcUK - Overclockers UK	12	243
5	Rechenkraft.net	10	236
6	TeAm AnandTech	8	60
7	BOINC@Poland	6	58
8	The Scottish Boinc Team	4	51
9	Czech National Team	2	31
10	Dutch Power Cows	1	31

 

[Fardringle]

Yup. Server processes are turned off, and he said he's not going to turn them back on for a while...

Project status

 

[Ken g6]

Well, that's one way of dealing with extra load on your BOINC server: shut down and wait hide until it goes away.

 

[Orange Kid]

Wanless Mersenne +2 Final
The 30 second sprint is over. Hardly a fair resolution, but it is what it is.
Between this sprint, the one simply cancelled and FB's poor showing of much interest in continuing to manage their site properly, I won't be putting much effort into keeping up with the sprints. I will do what I can for the rest of this season, but...
Also it seems a lot of projects are suddenly coming to an end. Don't know if there will be enough projects to last out the year.
Sorry to do this, but I guess all things must come to an end sooner or later.

1	UK BOINC Team	25	664
2	XtremeSystems	18	620
3	L'Alliance Francophone	15	397
4	OcUK - Overclockers UK	12	243
5	Rechenkraft.net	10	236
6	TeAm AnandTech	8	60
7	BOINC@Poland	6	58
8	The Scottish Boinc Team	4	51
9	Czech National Team	2	31
10	Dutch Power Cows	1	31

 

[StefanR5R]

Even though I did not take part: @Orange Kid, thanks for the stats and reports.

In light of the circumstance that the project owner is disinterested in a port of its application to the MS Windows operating system, it is, in my view, not overly surprising that he made weird decisions as project admin during this sprint. That's on top of the more fundamental problem that the science of this project is unsound, from what everybody else is saying.¹

I was however surprised that this project made it from an FB sprint candidate to an actual sprint project. @Ken g6, is it correct to assume that this happened because of one or more teams voting² for this one to be sprint project?

--------
¹) There is a bright side to the existence of the project: It is reminder to us to try and evaluate the validity of the projects that we donate computer capacity to.

²) If so, then foolishly, I dare say. Even regardless of the admin's actions during the sprint.

 

[Ken g6]

I was however surprised that this project made it from an FB sprint candidate to an actual sprint project. @Ken g6, is it correct to assume that this happened because of one or more teams voting² for this one to be sprint project?

I think so. I'm not sure; I lost track of voting for these projects for awhile. The site seemed to be having problems, and I just forgot.

The good news is I noticed the voting for the next sprint just before it closed. I voted for Rosetta.

 

[Orange Kid]

GPUGrid is up for the next sprint.
A tuff one to bunker.

 

[Fardringle]

I don't have much to add (just a GTX 1060 and a GTX 750ti), but the little that I do have will be working on GPUGrid during the sprint...

 

[StefanR5R]

From the 1st hourly stats update, it appears that TSBT managed to bunker well, while we were rather negligent this time. :-)

The 04:00 UTC announcement time was quite convenient for me as an inhabitant of the UTC+0200 time zone; but I admit that I was unprepared and started bunkering only more than half a day after this.

 

[TennesseeTony]

Ah man...

1.) The previous Sprint is now 'canceled' and we lost those points.

2.) My slippery mind forgot about the Sprint until now.

 

[Ken g6]

I got Breath Weapon working again yesterday. The power supply I ordered finally came in, but it didn't seem to help, until I noticed there was still one fan in the front of the case with a red LED. The fan wouldn't spin, and I think it may have a short. Unplugging it, everything else worked.

...Except that now I can't get fan control to work on the GPU ("Unknown error"), so it's in low-power mode.

 

[StefanR5R]

The previous Sprint is now 'canceled' and we lost those points.

The Formula BOINC owner multiplied all team points with the scientific value of WEP-M+2. :->

Sometimes it's nice to be able to pretend that something never happened.

 

[Ken g6]

...Except that now I can't get fan control to work on the GPU ("Unknown error"), so it's in low-power mode.

I got this solved too! https://hashcat.net/forum/thread-6036-post-34218.html#pid34218

 

[Fardringle]

I have a computer with dual Quadro 4000 GPUs and even though they are old, I decided to let them try to do some GPUGrid work for the sprint. However, the computer isn't getting any work from the project. Is the Quadro 4000 too old to run the current ACEMD tasks?

I don't get any errors, just "No tasks sent" in the BOINC event log when the computer asks for work...

 

[StefanR5R]

The computer with Quadro 4000s has got driver version 369.9 and Windows 10 x86-64.
For GPUGrid's Windows CUDA92 application, driver 396.26 or later is required; for the Windows CUDA101 application version, driver 410.48 or later. (source)

 

[StefanR5R]

@Fardringle,
the FAQ on required hardware is a badly edited mess. But parts of it hint that GPUGrid no longer works with Fermi cards, such as Quadro 4000.

 

[Fardringle]

The computer with Quadro 4000s has got driver version 369.9 and Windows 10 x86-64.
For GPUGrid's Windows CUDA92 application, driver 396.26 or later is required; for the Windows CUDA101 application version, driver 410.48 or later. (source)

That's the newest driver version available for the Quadro 4000, so I guess that's my answer. Thanks!

 

[Orange Kid]

GPUGrid day one

1	Gridcoin	25	73,554,265
2	The Scottish Boinc Team	18	32,476,416
3	XtremeSystems	15	31,266,079
4	UK BOINC Team	12	26,792,260
5	TeAm AnandTech	10	22,211,139
6	USA	8	20,402,905
7	Czech National Team	6	20,236,438
8	Rechenkraft.net	4	20,155,922
9	L'Alliance Francophone	2	9,910,056
10	SETI.Germany	1	9,153,058

 

[Orange Kid]

GPUGrid day two

1	Gridcoin	25	147,745,131
2	XtremeSystems	18	66,234,923
3	The Scottish Boinc Team	15	56,585,387
4	TeAm AnandTech	12	52,517,811
5	UK BOINC Team	10	52,166,367
6	USA	8	41,090,402
7	Czech National Team	6	40,820,014
8	Rechenkraft.net	4	40,719,494
9	L'Alliance Francophone	2	19,873,685
10	SETI.Germany	1	18,312,384

 

[Fardringle]

Passed the UK and gained quite a bit on the Scots. It was a good day!

 

[Ken g6]

Hey, @Orange Kid, whatever happened to those daily Free-DC listings? I liked seeing how much I contributed to the TeAm.

 

[StefanR5R]

From FreeDC:

	“yesterday”	“2 days ago”	total
xii5ku​	6,185,667​	6,787,127​	12,972,794​
10esseeTony​	5,051,597​	2,023,580​	7,075,177​
Howdy​	5,969,316​	511,864​	6,481,180​
biodoc​	3,333,790​	3,050,726​	6,384,516​
crashtech​	2,518,207​	1,980,444​	4,498,651​
emoga​	1,494,896​	1,509,353​	3,004,249​
Orange Kid​	1,413,456​	1,404,122​	2,817,578​
Lane42​	924,713​	980,913​	1,905,626​
phoenicis​	1,695,678​	0​	1,695,678​
far​	439,214​	735,574​	1,174,788​
Silverthorne​	257,533​	705,405​	962,938​
Fardringle​	238,968​	353,246​	592,214​
Ken_g6​	298,149​	248,825​	546,974​
Bremen​	112,530​	42,267​	154,797​
catavalon21​	22,097​	0​	22,097​
TurbineEngine​	0​	4,927​	4,927​
Σ​	29,955,811​	20,338,373​	50,294,184​