Formula BOINC in 2027?

Section where non-members/guests can read news about Formula BOINC
User avatar
UBT - Timbo
Site Admin
Posts: 1267
Joined: Wed Nov 23, 2016 1:21 am
Contact:

Formula BOINC in 2027?

Post by UBT - Timbo »

Hi all

There are just THREE more Sprints left this season (even if Formula 1 decide to relocate events due to take place in the Middle East - if they change the F1 race dates, then the relevant FB Sprint will also change dates).

And so we must start thinking ahead, about the 2027 FB Season and whether there should be any changes to FB.

My personal view is that we could consider:

a) Limiting the Sprints to only those weekends when an F1 Sprint is taking place - so for 2027, there would be 10 Sprints, instead of the current 24. This would solve the issue of trying to administer 24 Sprints, with an seemingly reduced number of projects that can actually sustain a full on Sprint, due to the bandwidth required to sustain multiple host machines downloading work units and uploading completed tasks within a limited time window.

b) Widen the number of Teams and/or Members in each League. Currently, I've tried to "group" both Teams and Members into Leagues where each group all have comparable capabilities in terms of credits earned per Sprint. This was done in the hope that those who have only a limited number of hosts can genuinely compete against other Teams and Members with similar capabilities. And hence FB points would be awarded for the efforts made by those Teams and Members who do NOT have a large farm of host machines. And so some healthy competition can take place in the lower Leagues.

c) Stop running FB (either the Sprints or both Marathon AND Sprints),due to fewer projects needing volunteers help, higher energy costs to run host machines and a general feeling that BOINC has had it's day, as many projects might be considered to be irrelevant or they are not publishing their results, so volunteers might be less interested in them.

d) Something else?

So, I'm open to ANY suggestions about how FB should run in 2027...the door is open for your opinions and suggestions.

best regards
Tim
UBT - wbiz
Registered FB member from 2023
Posts: 72
Joined: Fri Dec 16, 2022 4:03 pm

Re: Formula BOINC in 2027?

Post by UBT - wbiz »

With the current scoring scheme, the more members per league means the more people that get nothing for their efforts and the more that might simply give up sprints or give up half way through a sprint. I'd like like to see a bonus award for substantially taking part in a sprint. The fairest compromise I can think of is if the marathon boinc credits for a project are doubled during a sprint. This can be achieved fairly simply by adding the sprint credits to the marathon credits which doubles them up.

10 sprints sounds a bit small, less than one a month, I'd like to see them roughly 2 weeks apart, when you get three sprints on the run its very disruptive to the marathon projects and life in general. Could they be limited to a minimum of two weeks apart, I think it will have been 15 sprints in 2026, or a maybe max of two on the run which ends up with more (brain failed to count that). Somewhere in-between would be two weeks apart but slip one by a week where it would have been deleted immediately before a F1 break.

I'm sure everyone is aware that electricity prices in most countries are going to take a very significant hike again over the next few months and we have to expect another lull in boinc participation. We already have the large discrimination of cpu/gpu, there is another very significant discrimination between those with off-grid electricity with grid-only. Anything that crunches up the playing field would help but I understand that the point awards are not really up for negotiation due to the F1 association.
User avatar
UBT - Timbo
Site Admin
Posts: 1267
Joined: Wed Nov 23, 2016 1:21 am
Contact:

Re: Formula BOINC in 2027?

Post by UBT - Timbo »

UBT - wbiz wrote: Tue Nov 10, 2026 10:24 am With the current scoring scheme, the more members per league means the more people that get nothing for their efforts and the more that might simply give up sprints or give up half way through a sprint. I'd like like to see a bonus award for substantially taking part in a sprint. The fairest compromise I can think of is if the marathon boinc credits for a project are doubled during a sprint. This can be achieved fairly simply by adding the sprint credits to the marathon credits which doubles them up.
Hiya

Thanks for the feedback.

The basic premise of FB is that it takes place at the same time and day of each F1 event. So, to some degree, the entire annual Challenge just emulates what happens during an F1 season.

The number of Team and Member Leagues was increased, so that, as mentioned in the original post, more Teams and Members could earn FB points, even if they had limited resources at their disposal. And hence they would earn some "FB rewards" for their efforts, even in League 3.

The idea about increasing the number of Teams or Members in each league is to increase the competitiveness...whereas at present, virtually any Team who competes in Leagues 1 to 3 is bound to earn some FB points as each League is limited:

eg: Teams League 1 has 12 Teams, League 2 has 11 and League 3 has 9 participants

For the Members Leagues it's slightly different:

Members League 1 has 23 members, League 2 has 20 and League 3 has 26

Note that League 4 is reserved either for NEW Teams or Members and also for Teams/Members who appear to have retired from active participation.

These have all been placed in their respective Leagues based solely on their accumulated current BOINC credits earned on the Sprints (this season), over TWO time periods: 1) Total credits earned in the previous year and 2) Total credits earned after the first half of the current FB season.

The Member FB points awarded are tied into the F1 points awarded to each driver at the end of a race.

The Team FB points are awarded based on the entire Teams efforts during each Sprint...and on the ongoing Marathon League table positions, based on the running total of BOINC credits earned in the current year.

As things stand a Teams "Total FB points" is already the sum of ongoing "Marathon FB points" added to the "Total FB Sprint points" earned over every Sprint they have taken part in.

As per here:

https://www.formulaboinc.org/index.py?r ... nding=team

So, I just wanted to get some views on whether anything should change on these principles.
UBT - wbiz wrote: Tue Nov 10, 2026 10:24 am 10 sprints sounds a bit small, less than one a month, I'd like to see them roughly 2 weeks apart, when you get three sprints on the run its very disruptive to the marathon projects and life in general. Could they be limited to a minimum of two weeks apart, I think it will have been 15 sprints in 2026, or a maybe max of two on the run which ends up with more (brain failed to count that). Somewhere in-between would be two weeks apart but slip one by a week where it would have been deleted immediately before a F1 break.
The timing of FB Sprints is purely down to when F1 hold each of their events...so there have been in the past, 18, 20, 22 F1 races and the FB Sprint events have been taking place on SOME of the same weekends as the races.

So, we are directly "controlled" by whenever F1 holds their F1 races, so this cannot be changed.

And in recent years, F1 has adopted Sprint races on the day before a full race event and for 2027, there will be an increase in the number of F1 Sprints to 10 Sprints, (where in previous seasons, there were just a few).

And as mentioned, given the limited number of project who can sustain a FB Sprint event, so it just seems a bit more sensible to make the FB Sprints more "special" to only hold 10 FB Sprints, rather than trying to support 24 FB Sprints from only maybe 6 or 8 suitable projects.

It boils down to whether Teams/Members want to be crunching the same project (over various FB Sprints) multiple times per season. Limiting the Sprints to 10, would ensure that most FB Sprints would feature different projects each time.
UBT - wbiz wrote: Tue Nov 10, 2026 10:24 am I'm sure everyone is aware that electricity prices in most countries are going to take a very significant hike again over the next few months and we have to expect another lull in boinc participation. We already have the large discrimination of cpu/gpu, there is another very significant discrimination between those with off-grid electricity with grid-only. Anything that crunches up the playing field would help but I understand that the point awards are not really up for negotiation due to the F1 association.
Yes, I'd think you've made a good point. But it sort of re-enforces the idea to limit the number of Sprints to 10, instead of having 24. Then members can concentrate more on the Marathon projects they prefer, and to then save money by not having to take part in 24 high-energy consuming FB Sprints.

Sadly, there is no easy way to differentiate between hosts that are part of a private home network and any BOINC credits earned by "cloud based" resources...and of course between those who use their own electricity (provided by an energy supplier) and any power generated by solar/wind locally. And I doubt even the projects have such information (except by monitoring the IP addresses to see which hosts are based on cloud based servers).

regards
Tim
Antares2022
Registered FB member from 2023
Posts: 15
Joined: Sun Apr 02, 2023 12:44 pm

Re: Formula BOINC in 2027?

Post by Antares2022 »

Kessler@home would be, at the very least, a marathon project. 51 teams are currently active.

Regards,
Antares2022
User avatar
UBT - Timbo
Site Admin
Posts: 1267
Joined: Wed Nov 23, 2016 1:21 am
Contact:

Re: Formula BOINC in 2027?

Post by UBT - Timbo »

Antares2022 wrote: Mon Nov 16, 2026 5:15 am Kessler@home would be, at the very least, a marathon project. 51 teams are currently active.

Regards,
Antares2022
Hi

Thanks for the feedback.

I've noticed a few new BOINC projects have started recently, but as with some other well-known projects, one must ask the question as to:
a) what server hardware do they have and how good is their internet connectivity?
and
b) does the project have worthwhile aims that Teams and Members feel happy about supporting?

I'm not going to question *any* project as far as their motivation goes, as I was one of the crunchers who earned a lot of BOINC credits crunching "tasks" for BitcoinUtopia some years back :oops: This generated a huge boost to my overall total BOINC credits, but that project was not created to solve any real world science or mathematics problems...it was just a credits-fest, if you had some spare cash to buy a Bitcoin miner and the required power supplies and associated running costs to operate them :o

There is still time (about 6 weeks) for us to assess which projects could be added to the Marathon and the same can be said for the Sprints too. So any suggestions would be welcome.

regards
Tim
Antares2022
Registered FB member from 2023
Posts: 15
Joined: Sun Apr 02, 2023 12:44 pm

Re: Formula BOINC in 2027?

Post by Antares2022 »

Hi

Kessler@home is a NASA project.
I don’t think this will be one of those short-lived projects.

Regards

Antares2022
User avatar
Tamagoch
Registered FB member from 2023
Posts: 68
Joined: Sat Jul 22, 2023 11:58 am
Location: Germany

Re: Formula BOINC in 2027?

Post by Tamagoch »

Antares2022 wrote: Mon Nov 16, 2026 6:07 pm Kessler@home is a NASA project.
I don’t think this will be one of those short-lived projects.
Is it not NASA official, I believe.
The project has already hit 1% of total work done. Unless it is planned to extend the search somehow, it may be finished earlier than expected.

As Tim said, we have 6 weeks to observe these new project and make some decisions.
User avatar
UBT - Timbo
Site Admin
Posts: 1267
Joined: Wed Nov 23, 2016 1:21 am
Contact:

Re: Formula BOINC in 2027?

Post by UBT - Timbo »

Antares2022 wrote: Mon Nov 16, 2026 6:07 pm Hi

Kessler@home is a NASA project.
I don’t think this will be one of those short-lived projects.

Regards

Antares2022
Hi

Looking at the Kessler "about" page here:

https://kesslerathome.org/about.php

it says:
Kessler@home is an independent project. We aim to work with established debris-modeling groups and will credit collaborators here as the project develops.
So, I presume, that the project collects freely available data about the orbits of many satellites and other "objects" that are left-overs from previous space launches, and then it is using volunteer host machines to help "predict" any possible debris fields due to in-orbit collisions taking place if (say) "Object A" collided with "Satellite B".

And the resulting "domino effect" if the resulting debris then collided with other satellites in stable orbits, resulting in even more debris being created, which then collides with even more satellites - hence the possibility of the Kessler Syndrome taking place, where nothing can be launched from Earth as every orbit is full of very, very fast moving small particles of debris that could hit any spacecraft hoping to reach space.

regards
Tim
User avatar
UBT - Timbo
Site Admin
Posts: 1267
Joined: Wed Nov 23, 2016 1:21 am
Contact:

Re: Formula BOINC in 2027?

Post by UBT - Timbo »

Tamagoch wrote: Mon Nov 16, 2026 9:02 pm Is it not NASA official, I believe.
The project has already hit 1% of total work done. Unless it is planned to extend the search somehow, it may be finished earlier than expected.

As Tim said, we have 6 weeks to observe these new project and make some decisions.
Hiya

Yes, the Kessler website says:
The main sweep started on 6 November 2026. The work generator takes the 302,400 scenarios one after another, in a fixed random order, and sends out 400 workunits of each, one per seed, so results arrive scenario by scenario, and the scenarios finished at any time are a random sample of the whole map.
And from their homespage:

https://kesslerathome.org/

as of today they have completed 1.139% out of the 302,400 scenario's.

If we assume they started on 6th November, then after 12 days they have completed 3,964 scenarios out of the 302,400 (let's say "1.2%").

So, 1.2% over 12 days implies 0.1% per day, so 100% might be achieved in about 1,000 days*.

(*This assumes that the number of hosts does NOT increase. Less time will be needed if the number of hosts increases).

So, it is possibly suitable as a Marathon project.

regards
Tim
ChelseaOilman
Registered FB member from 2024
Posts: 3
Joined: Sun Sep 15, 2019 10:22 pm

Re: Formula BOINC in 2027?

Post by ChelseaOilman »

Ian&Steve C. — 9:42 AM
the first sweep is an exhaustive search of all the possible combinations of variables for scenarios (11 variables swept). that amounts to 302,400 scenarios. 400 WUs (seeds) per scenario, at least 2 tasks per WU (for validation) = 240 million + tasks.

the first sweep is designed to weed out the borderline and more uncertain or interesting scenarios. which will be re-run with even more seeds per scenario to get a closer look at them. something like 2000 seeds on a smaller number of scenarios.
i honestly want to get enough power to get both sweeps done inside a year. the faster the better. because we want to get the results to write our papers. I don't think projects should be "never ending" so to speak.

but I do have some loose avenues that I can explore for more future work to keep the project interesting and going. So far things are going pretty smoothly, and I have some contacts with others in the industry, and I'll probably give a presentation about our project at the next SSA Operators conference in Denver if the organizer thinks it's worth sharing. My hope is that someone would see our platform as proven and be interested to use us for hosting more/similar research

(edited by admin 21st Nov 2026 - it was about restoring/deleting another users forum account that could not be accessed - now fixed)
Post Reply