2019 Formula Boinc Sprints

[crashtech]

I wonder why my RTX 2060 looks worse than a 1070 in Einstein, but better than a 1070ti in F@H?

 

[ao_ika_red]

I remember Einstein was quite late with their Pascal support, it's probably the same with Turing. But it doesn't make sense because it's almost a year already.

 

[zzuupp]

I remember Einstein was quite late with their Pascal support, it's probably the same with Turing. But it doesn't make sense because it's almost a year already.

IIRC, they're justifiably paranoid about: internal testing, and being absolutely correct always.

I'll add a dash of speculation about now only having basic manufacturer support without any subsidies.

 

[crashtech]

I guess I'll be hanging to the few HD 7970's I have left, since most complete Einstein WUs in well under 10 minutes.

 

[biodoc]

@StefanR5R usually provides last year's Sprint performance data from time to time but since he's traveling, I'll give it a go. We took 1st place last year in league 1 but [H] in league 2 beat us by a large margin. Gridcoin will be a big factor this year because they are averaging 31 M PPD so their total this year will be `93 M points which is more than we produced last year.

Spoiler: 2018 Einstein sprint results

League 1:

1 TeAm AnandTech 85696216
2 Gridcoin 54195892
3 SETI.USA 49584831
4 BOINC @ AUSTRALIA 46263266
5 SETI.Germany 35161845
6 Czech National Team 31632776
7 Rechenkraft.net 28981258
8 UK BOINC Team 28851652
9 Russia Team 27640531
10 Francophone Alliance 24271565

League 2:

1 [H] ard | OCP 121196097
2 XtremeSystems 80855447
3 Crunching @ EVGA 38084191

Combined:

1 [H] ard | OCP 121196097
2 TeAm AnandTech 85696216
3 XtremeSystems 80855447
4 Gridcoin 54195892
5 SETI.USA 49584831
6 BOINC @ AUSTRALIA 46263266
7 Crunching @ EVGA 38084191
8 SETI.Germany 35161845
9 Czech National Team 31632776
10 Rechenkraft.net 28981258
11 UK BOINC Team 28851652
12 Russia Team 27640531
13 Francophone Alliance 24271565

 

[Ken g6]

I wanted to add our Pentathlon performance here, but there were so many team alliances going on that a comparison is [H]ard to make.

 

[IEC]

I opened the floodgates, but the servers are only accepting a trickle at a time. Oh well.

 

[emoga]

I opened the floodgates, but the servers are only accepting a trickle at a time. Oh well.

Do you have the <max_file_xfers> tag in your cc_config?

You can specify how many download/upload slots can be active at one time.

Since I have good internet I have it at:

<max_file_xfers>100</max_file_xfers>
<max_file_xfers_per_project>100</max_file_xfers_per_project>

10 or 20 might be more reasonable though.

Spoiler: Example cc_config.xml

<cc_config>
    <options>
    <allow_multiple_clients>1</allow_multiple_clients>
    <dont_suspend_nci>1</dont_suspend_nci>
    <report_results_immediately>1</report_results_immediately>
    <allow_remote_gui_rpc>1</allow_remote_gui_rpc>
    <use_all_gpus>1</use_all_gpus>
    <max_file_xfers>20</max_file_xfers>
    <max_file_xfers_per_project>10</max_file_xfers_per_project>
    </options>
</cc_config>

 

[zzuupp]

I just droppped a baby bunker: 12 hours on one GPU, and some CPU tasks.

 

[crashtech]

Interestingly enough, the iGPU in my 8700K seems to be able to run Einstein passably well.

Edit: Oh, they are totally different WUs, of course.

 

[crashtech]

For anyone running Einstein CPU tasks, the "Continuous Gravitational Wave search O2 All-Sky" tasks are pretty long, so depending on your CPU, you may want to abort ones that haven't started by tomorrow sometime. I'm proactively getting rid of them now.

 

[ao_ika_red]

I did that 12 hours ago. It eats quite a lot of RAM too.

 

[IEC]

The lads from Oz are hot on our tails. It's almost like it's winter and they're taking advantage of the cold or somethin'

 

[Orange Kid]

Einstein@home day one

We are in fourth at the moment but have been trading places with [H] off and on. The Aussies are on our heels too. Promises to be and exciting one.



From FreeDC

 

[zzuupp]

I don't know about y'all, but I have huge numbers of PENDING. Alas, invalid vs valid is a tie at two dozen. The red card pukes on one every hour or so.

 

[crashtech]

Something is definitely going on, hopefully it gets resolved, but I've been seeing "down for maintenance" messages for many hours now.

 

[biodoc]

The server seems to be functioning normally the last few hours. Hopefully that will continue.

 

[Orange Kid]

Einstein@home day two.
Still trading places with [H], but at this moment we are in fourth.
The pending tasks seem to be still piling up. Will be fun when the dumps begin.



From FreeDC

 

[TennesseeTony]

I can go all in for only about 12 hours....night time here and the air conditioning can keep up now....almost....

Thanks Kid for all that you do!

edit: I miss Stefan. :/

 

[zzuupp]

I'm now sitting on 149 pending / 66 valid / 44 oops.

Thanks for the stats, OK!

 

[TennesseeTony]

Lots of computation errors in this batch of tasks...

 

[Howdy]

Well unfortunately it looks like we'll be fighting to keep 3rd place against GC. Although SUSA has a smallish lead (pending they drop something) it seems to be enough to keep 2nd out of reach.

 

[TennesseeTony]

This one is quite the nail-biter! We have 2nd at the moment, but......

 

[TennesseeTony]

I have plenty, grab yourself some!

 

[pututu]

Congratz on finishing 2nd. What a super nail biter!