Please visit donation page to help the project cover running costs for this month

Toggle Menu

Join PrimeGrid

Returning Participants


Leader Boards



11) Message boards : Proth Prime Search : Multithreading PPSE (Message 125347)
Posted 322 days ago by Profile pschoeferProject donor
I would guess that the gap depends heavily on how many active users there are, because the second task has to be send to a different user. When a large contributer fetches lots of tasks, it may take a while until sufficiently many other users have requested tasks. This effect probably also worked to dthonon's advantage last year.
12) Message boards : Number crunching : 2019 Challenge Voting (Message 123713)
Posted 359 days ago by Profile pschoeferProject donor
Good job. Many different FFT lengths, not too long in total, although I'd like to see at least one of the short task challenges (PPS, AP27, GFN-17-Low) running only for 3 days. Two comments regarding the themes:

1) I'm not sure about the 25th anniversary of discovery of the first extrasolar planets. Wolszczan & Frail reported two planets around a pulsar in January 1992, and the first discovery around a main-sequence star was announced in October 1995 (Mayor & Queloz).

2) While I understand the problem with 'Winter Solstice', there are still two solstices per year. 'December Solstice' would be a neutral, but unambiguous name. :)
13) Message boards : General discussion : Fiddling with the computer because I can't sleep (Message 122858)
Posted 383 days ago by Profile pschoeferProject donor
How did you check the stability? Your computer returned several wrong results (the invalid and, very likely, the inconclusive tasks) last week, so it was not stable before.

Running those long LLR tasks can be even more demanding than a Prime95 torture test (depending on the settings). That is, the latest version of Prime95, not the ancient pre-AVX version that some overclockers are recommending because they only care about stability for games.
14) Message boards : Number crunching : 2019 Challenge Voting (Message 122661)
Posted 388 days ago by Profile pschoeferProject donor
1. SoB - it was already mentioned somewhere that there will very likely be a challenge every year until the DC is finished, so we won't get around that one anyways. :)

2. ESP - Lennart Honary Challenge is great, but a 65 day challenge would only be special in the sense that it would become especially boring long before it is finished. How about one challenge of two times 6.5 days with a break in between (could be a 52 days break so that it is still 65 days from start of part 1 to end of part 2)?

3. TRP - there should be a third "conjecture challenge" and it should be shorter than 10 days if we already have SoB and ESP challenges in the 10+ days range, so TRP is the more reasonable choice than PSP.

4. GCW - we had both CUL and WOO this year, let's do the combination of their generalized forms next year (ignoring the fact that it is actually GC only ;)).

5. PPSE - we had MEGA in 2016 and plain PPS challenges every other year, but no PPSE since 2010. Ideally suited for a sprint of 1 or 2 days.

6. GFN - we had the short ones (15-17) last year and a long one (21) this year, so it's time for something in between again (18, 19, or both).

7. AP27 - another GPU project for the traditional Winter Solstice challenge and AP27 is more interesting than PPS Sieve.

There is a Conjunction of Venus and Jupiter on November 24th next year, so there could be a 5 day challenge (GCW or medium-sized GFN) from Nov 24 to Nov 29 with that theme which did not fit this year.
15) Message boards : Problems and Help : Team Country change (Message 122457)
Posted 393 days ago by Profile pschoeferProject donor
Not sure if PrimeGrid imports the BOINC-wide team database, but Storm exists in that database with country set to Ireland. Maybe that's the right place to change it.
16) Message boards : Number crunching : New PrimeGrid Policy - Monetization of BOINC credit (Message 121868)
Posted 403 days ago by Profile pschoeferProject donor
I totally agree with everything written in this thread so far.
17) Message boards : General discussion : Member Poll: Is Gridcoin good or bad? (Message 121669)
Posted 407 days ago by Profile pschoeferProject donor
Cryptocurrencies in general are a very polarizing topic, so it is probably a wise idea for any project to keep some distance. If they want to use a project to determine how many of their "coins" each participant receives, then they have to figure out how to do this without DDOSing the project's servers. If they can't come up with a solution, well, that's their problem, not the project's.

Of course, individual users should be banned if they are caught cheating. Yes, there have been cheaters and even another botnet all the time even without any kind of monetary incentive, that's why the quorum for sieve tasks was set to 2 before Gridcoin was even created. But unless their aim was to just create a mess by injecting fake results (and at least that botnet did not do that), their success was not that great, as their points were subtracted later and their reputation was ruined. If they now cheat to receive more coins, they won't care about having their points subtracted later. While members of the Gridcoin community regularly helped the project administrators when something fishy was going on in the past, I wonder why they did not temporarily remove all projects that are open to exactly the same attack from their whitelist to let them fix the loophole before it is exploited. And the cheaters caught so far were not particularly clever; heaven help us if those coins ever have any real value and more skilled hackers have a go at it.

Even apart from that, I have trouble to find positive aspects of Gridcoin. Mike already noted that he does not see a net positive for PrimeGrid. While PrimeGrid always has been my pet project, I've also been looking at how the BOINC platform as a whole develops. It's perhaps not the most useful metric with all the flaws of cross-project user identification, but the rise of Gridcoin has not stopped the decline of active crunchers. I'm not aware of any numbers on how many new crunchers they recruited, but at this year's BOINC workshop, a Gridcoin member claimed that their most successful recruitment campaign was an advertising campaign at BOINCstats. It's a safe bet that every visitor of BOINCstats already is an active cruncher, so that's a grand total of zero new crunchers. A significant portion of their computing power is coming from the Charity Engine, which has been in operation long before (albeit far less visibly unless you were looking for it very carefully). The fraction of active members in team Gridcoin is somewhat higher than in other big teams here at PG, but of course, the other big teams had much more time for members to quit crunching. On younger projects, there is no difference, so the monetary incentive doesn't appear to keep crunchers more motivated in the long term. That's not yet evidence of monetary incentives killing enthusiasm as some studies on that topic suggest, but the monetary incentive also doesn't appear to be that effective an argument to recruit more crunchers.

Edit: Wow, I really need to write faster.
18) Message boards : Number crunching : GTX 1060 with a GTX 1070? (Message 121151)
Posted 421 days ago by Profile pschoeferProject donor
It will use both if you use that cc_config.xml.
19) Message boards : Number crunching : GTX 1060 with a GTX 1070? (Message 121144)
Posted 421 days ago by Profile pschoeferProject donor
Most likely, you will need the following cc_config.xml file in your BOINC data directory (or add the use_all_gpus line if you already have this file):

<cc_config> <options> <use_all_gpus>1</use_all_gpus> </options> </cc_config>

With default settings, BOINC uses only the "most capable" GPU(s), i.e. the 1070 in your case. Apart from this, you shouldn't run into Problems, as your GPUs are sufficiently similar.
20) Message boards : News : Genefer 3.3.4 now live (Message 121063)
Posted 425 days ago by Profile pschoeferProject donor
Have you ever downloaded a GFN-21 task on that machine? If not, everything is working as intended and the message will disappear once you do that.

Next 10 posts
[Return to PrimeGrid main page]
Copyright © 2005 - 2019 Rytis Slatkevičius (contact) and PrimeGrid community. Server load 0.76, 0.94, 1.10
Generated 12 Dec 2019 | 19:14:56 UTC