Join PrimeGrid
Returning Participants
Community
Leader Boards
Results
Other
drummers-lowrise
|
Message boards :
Seventeen or Bust :
The SOB Double Check will end... aka The Way Too Early Prediction Thread
| Author |
Message |
|---|
 Jay Send message
Joined: 27 Feb 10
Posts: 132
ID: 56067
Credit: 64,433,752
RAC: 17,531
                    
|
|
Based on the rate so far (from Michael's x% complete posts) and wild predictions regarding the impact of future challenges and increasing technological capabilities, I'm going to predict the SOB Double Check will be complete on May 28, 2023.
Anyone else? What date are you predicting? | |
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 14011
ID: 53948
Credit: 433,157,676
RAC: 1,017,804
                              
|
|
It will be interesting to see which happens first: we finish the DC or Elon Musk successfully lands something on Mars. Not a person, but any sort of lander.
As for predictions, I think you're in the right ballpark.
____________
My lucky number is 75898524288+1 | |
|
|
|
|
In order to help eliminate the computing time cost of data loss in the future, is there any chance that the complete result set be made available periodically for download? That way we'll have tons of users with the data to restore from in case anything ever does happen.
____________
 | |
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 14011
ID: 53948
Credit: 433,157,676
RAC: 1,017,804
|
In order to help eliminate the computing time cost of data loss in the future, is there any chance that the complete result set be made available periodically for download? That way we'll have tons of users with the data to restore from in case anything ever does happen.
That's not a terrible idea.
Understand that both Jim and myself are really paranoid about business continuity, and it's hard to imagine any sort of disaster that would result in the data being lost. I can think of things that would shut PrimeGrid down permanently, but the data would survive such an event. I might need to use sneakernet to distribute the data (the proverbial station wagon filled with 9-track tapes), but short of a nuclear holocaust hitting multiple locations, the data is well preserved. And if there's s single event that takes out both the data center and the secondary storage site, which is hundreds of miles away, I suspect what's left of civilization won't be worrying about our data.
When 9/11 happened, I was working for a large company that had a substantial data center INSIDE one of the World Trade Center towers. The data center itself was destroyed, of course. All of our people working in that data center survived (it was on the 3rd floor of a 110 story building.) Our business didn't miss a beat. The tower came down, those computers stopped communicating, and our computers in other datacenters simply picked up the workload. PrimeGrid's disaster recovery plan isn't nearly as robust, but we won't have one of those "oh God, everything is lost" type of disasters. Worst case scenario, if there's still power and computers, we'll be able to get PrimeGrid up and running again, sooner or later, with minimal loss of data.
But it's not a terrible idea, and we'll give it some thought.
____________
My lucky number is 75898524288+1 | |
|
|
|
|
Ever the optimist of enhanced CPU instruction extensions, I will say we'll finish February 2, 2022
I'm a fan of the sneakernet backup, I keep an encrypted backup drive in my work locker and someday I'll have a safety deposit box and leave one in there, too. How big is the Primegrid results database?
____________
Eating more cheese on Thursdays. | |
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 14011
ID: 53948
Credit: 433,157,676
RAC: 1,017,804
|
Ever the optimist of enhanced CPU instruction extensions, I will say we'll finish February 2, 2022
I'm a fan of the sneakernet backup, I keep an encrypted backup drive in my work locker and someday I'll have a safety deposit box and leave one in there, too. How big is the Primegrid results database?
It's not that simple to answer, because there's lots of archived data that isn't in the database. An easier question is "How much disk space is used?" Less than 250 GB for BOINC. Less (a lot less) for PRPNet. Much of that is compressed. We could fit the whole thing on a thumb drive if we wanted.
As I said before, there's currently little risk of the data being lost. Almost everything exists in at least 4 locations -- 2 servers, an incremental cloud backup, and an offsite storage. Those backups are versioned, meaning there's multiple copies of each file. Even an encryption/malware attack would be survivable. Datacenter burns to the ground? Survivable. Rackspace folds up shop without warning (or we get locked out)? Survivable. Nuclear attack? Maybe survivable, maybe not. Depends on where the nukes fall. It will take at least two nukes in the right locations.
____________
My lucky number is 75898524288+1 | |
|
MonkeydeeVolunteer tester
Send message
Joined: 8 Dec 13
Posts: 540
ID: 284516
Credit: 1,529,047,472
RAC: 770,698
                    
|
Based on the rate so far (from Michael's x% complete posts) and wild predictions regarding the impact of future challenges and increasing technological capabilities, I'm going to predict the SOB Double Check will be complete on May 28, 2023.
Anyone else? What date are you predicting?
Based upon the way PSP progressed (I know it's apples to oranges here) I would say that we would have all doublecheck units sent out for the first time by December 1, 2019. However, given how SoB works with wing-people and resends due to various task failures or abandonments I would say that it would take until April 1, 2021 to get all the results back.
Of course all of these predictions will be thrown out the window if someone finds a prime.
____________
My Primes
Badge Score: 4*2 + 6*2 + 7*4 + 8*9 + 11*3 + 12*1 = 165
 | |
|
|
|
|
Meanwhile (probably my understanding is lacking here) it just happened to me: is The Prime Rank on the home correct for SOB? Just a consistency question, you know.
____________
I'm counting for science,
Points just make me sick. | |
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 14011
ID: 53948
Credit: 433,157,676
RAC: 1,017,804
|
Meanwhile (probably my understanding is lacking here) it just happened to me: is The Prime Rank on the home correct for SOB? Just a consistency question, you know.
No.
It's still picking up the straggler in-progress tasks. Until those complete, the largest numbers still be worked on are technically leading edge tasks.
____________
My lucky number is 75898524288+1 | |
|
|
|
|
Well, that's what I've expected. I understand that making it dynamic is too much coding (besides being useless). However, for sake of integrity, footnote it maybe?
____________
I'm counting for science,
Points just make me sick. | |
|
GellyVolunteer tester
Send message
Joined: 13 Nov 16
Posts: 49
ID: 468732
Credit: 2,397,279,885
RAC: 0
          
|
|
Considering the insanity that was the SoB challenge this January, and assuming (and hoping and praying) that it's done every year until it's done, my basic graphing skills and line drawing points me to April 22, 2019 .
It's pretty amazing that the 11 or so percentage points that were double checked in half a month were almost equivalent to the half a year the previous 11 took. And it seems like the rate of SoB double check is slightly higher now, even after the challenge and during a TdP. | |
|
|
|
|
November 2020
____________
(252^6548-1)^2-2 is prime! Small, but mine.
134137784^32768+1(DC)
107853608^8192+1(DC)
10465966^16384+1(DC)
 | |
|
|
|
|
I agree that chalenge makes it quicker. So I think: June 4, 2019. | |
|
dthononVolunteer tester
Send message
Joined: 6 Dec 17
Posts: 435
ID: 957147
Credit: 1,749,537,640
RAC: 324,050
                      
|
|
I would not be so optimistic. I have been processing around 1000 SOB per month since early December. But that will end this month as we decommission our VMs. I don't know what part of the monthly SOB processing that means, but it should be significant. Better wait a couple of months to see what the rate is at that time.
| |
|
HonzaVolunteer moderator
Volunteer tester
Project scientist
Send message
Joined: 15 Aug 05
Posts: 1957
ID: 352
Credit: 6,137,993,476
RAC: 2,260,339
                           
|
... I have been processing around 1000 SOB per month since early December. But that will end this month as we decommission our VMs. I don't know what part of the monthly SOB processing that means, but it should be significant...
http://www.primegrid.com/server_status_tasks.php
Note that February had very low performance as many participants were focusing on PPS/PPSE/PPSMega.
And, first half of January had extra performance due to SoB challenge.
____________
My stats | |
|
robishVolunteer moderator
Volunteer tester
Send message
Joined: 7 Jan 12
Posts: 2210
ID: 126266
Credit: 7,513,406,158
RAC: 3,355,962
           
|
|
September 2019
____________
My lucky numbers 10590941048576+1 and 224584605939537911+81292139*23#*n for n=0..26 | |
|
AndySend message
Joined: 27 Nov 17
Posts: 24
ID: 953480
Credit: 180,848,254
RAC: 55,758
  
|
|
I've thrown my roughly 50 cores (Of widely varying quality) on the fire, but we haven't gone from 30% -> 31% in nearly 2 weeks, So best-case 24% per year from 'normal' operation, and a 13% or so from a challenge, gives us pretty close to 2 years from now, assuming 2 more challenges in those time frames.
I'm actually going to throw down a really shady guess and say on the 9th day of the 2020 SOB challenge. | |
|
|
|
|
Will the last unfinished WUs be finished manually? The finish date depends on it. | |
|
|
|
|
July 23rd, 2019 | |
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 14011
ID: 53948
Credit: 433,157,676
RAC: 1,017,804
|
Will the last unfinished WUs be finished manually? The finish date depends on it.
Probably.
There's 2 different finish dates: When the last DC tasks are initially sent out, and we start working once again on real leading edge candidates, and when (probably 3 to 6 months later) the wingmen finally finish working on the DC tasks.
We are down to 5083 candidates that have not yet been turned into tasks, out of a total of 149364 candidates that needed to be checked. We're burning through about 100 each day, give or take.
____________
My lucky number is 75898524288+1 | |
|
JaySend message
Joined: 27 Feb 10
Posts: 132
ID: 56067
Credit: 64,433,752
RAC: 17,531
|
September 2019
A super quick perusal of the responses looks like this one is the closest!
Congratulations! You win, uh, some brief mention of your successful prediction...
Great job to everyone who helped on the double check. Lets keep at it and finish the conjecture! | |
|
dthononVolunteer tester
Send message
Joined: 6 Dec 17
Posts: 435
ID: 957147
Credit: 1,749,537,640
RAC: 324,050
|
September 2019
A super quick perusal of the responses looks like this one is the closest!
Congratulations! You win, uh, some brief mention of your successful prediction...
Great job to everyone who helped on the double check. Lets keep at it and finish the conjecture!
Is it a precise prediction, or another outstanding piece of luck ? Seeing who proposed this date, both are probably true ;-)
| |
|
|
|
It will be interesting to see which happens first: we finish the DC or Elon Musk successfully lands something on Mars. Not a person, but any sort of lander.
As for predictions, I think you're in the right ballpark.
Was looking through old posts, and of course we finished it earlier than Elon Musk :P
Also, Understand that both Jim and myself are really paranoid about business continuity, and it's hard to imagine any sort of disaster that would result in the data being lost. I can think of things that would shut PrimeGrid down permanently, but the data would survive such an event. I might need to use sneakernet to distribute the data (the proverbial station wagon filled with 9-track tapes), but short of a nuclear holocaust hitting multiple locations, the data is well preserved. And if there's s single event that takes out both the data center and the secondary storage site, which is hundreds of miles away, I suspect what's left of civilization won't be worrying about our data.
What would shut down PG permanently? Just asking.
____________
My lucky number is 6219*2^3374198+1
 | |
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 14011
ID: 53948
Credit: 433,157,676
RAC: 1,017,804
|
It will be interesting to see which happens first: we finish the DC or Elon Musk successfully lands something on Mars. Not a person, but any sort of lander.
As for predictions, I think you're in the right ballpark.
Was looking through old posts, and of course we finished it earlier than Elon Musk :P
Also, Understand that both Jim and myself are really paranoid about business continuity, and it's hard to imagine any sort of disaster that would result in the data being lost. I can think of things that would shut PrimeGrid down permanently, but the data would survive such an event. I might need to use sneakernet to distribute the data (the proverbial station wagon filled with 9-track tapes), but short of a nuclear holocaust hitting multiple locations, the data is well preserved. And if there's s single event that takes out both the data center and the secondary storage site, which is hundreds of miles away, I suspect what's left of civilization won't be worrying about our data.
What would shut down PG permanently? Just asking.
Rytis dies of Covid-19? Everyone gets bored and stops crunching? All the admins get tired and nobody else wants to do it? World War III? Economic disruption from the pandemic is so bad that we can't afford to keep the project running?
The USUAL things that kill a project are either "disk crashed and I lost all the data", or "server died and I can't afford to replace it." Those are not things most hobbyists plan for (or even most University-run projects). PrimeGrid's pretty well protected against that kind of failure.
____________
My lucky number is 75898524288+1 | |
|
|
|
It will be interesting to see which happens first: we finish the DC or Elon Musk successfully lands something on Mars. Not a person, but any sort of lander.
As for predictions, I think you're in the right ballpark.
Was looking through old posts, and of course we finished it earlier than Elon Musk :P
Also, Understand that both Jim and myself are really paranoid about business continuity, and it's hard to imagine any sort of disaster that would result in the data being lost. I can think of things that would shut PrimeGrid down permanently, but the data would survive such an event. I might need to use sneakernet to distribute the data (the proverbial station wagon filled with 9-track tapes), but short of a nuclear holocaust hitting multiple locations, the data is well preserved. And if there's s single event that takes out both the data center and the secondary storage site, which is hundreds of miles away, I suspect what's left of civilization won't be worrying about our data.
What would shut down PG permanently? Just asking.
Rytis dies of Covid-19? Everyone gets bored and stops crunching? All the admins get tired and nobody else wants to do it? World War III? Economic disruption from the pandemic is so bad that we can't afford to keep the project running?
The USUAL things that kill a project are either "disk crashed and I lost all the data", or "server died and I can't afford to replace it." Those are not things most hobbyists plan for (or even most University-run projects). PrimeGrid's pretty well protected against that kind of failure.
Huge oof on the second and fourth. 🤣🤣🤣
Also, isn't the fifth a temporary shutdown?
I never knew BOINC projects were so "vulnerable" 😂
____________
My lucky number is 6219*2^3374198+1
| |
|
Dave
Send message
Joined: 13 Feb 12
Posts: 3207
ID: 130544
Credit: 2,285,331,764
RAC: 787,864
           
|
|
Nothing is forever. We learn this in life. | |
|
KEPSend message
Joined: 10 Aug 05
Posts: 301
ID: 110
Credit: 12,352,853
RAC: 139
      
|
I never knew BOINC projects were so "vulnerable" 😂
They are :( ... a classic example of a project disappering in a blink of an eye, is the former RieselSieve.com project... It took quite a while, to convince the admins back then, that RieselSieve was gone and PG could/should overtake the effort. No one has revealed what happened, except a few different conspiracy theories, like CIA or FBI shutting down RieselSieve, but no proves has been delivered to support that theory. Maybe they just found a MegaPrime and decided to quit or they ran out of means to continue the effort - who knows. They sure did vanish from one moment to the next. Thankfully PG eventually overtook the conjecture and work resumed for TRP :) | |
|
|
|
|
Oh yes I remember RieselSieve. I donated funds to the project admin, B2, to buy some RAM sticks then the project blocked internet traffic from Australia so my world rank dropped. Then the project just went dark. Not impressed. | |
|
BurVolunteer tester
Send message
Joined: 25 Feb 20
Posts: 515
ID: 1241833
Credit: 414,481,880
RAC: 219
              
|
|
The doubleckecking probably is in regard to work already done by Rieselsieve? Are these doublechecks doublechecked at PG as well? Or are these special tasks that are only run once and if same result as Rieselsieve they're considered authorative?
____________
1281979 * 2^485014 + 1 is prime ... no further hits up to: n = 5,700,000 | |
|
|
|
The doubleckecking probably is in regard to work already done by Rieselsieve? Are these doublechecks doublechecked at PG as well? Or are these special tasks that are only run once and if same result as Rieselsieve they're considered authorative?
It was a double-check of salvaged work residues that had not yet been checked (or were not recorded as checked) during the original RS time. There were also plenty of previously completed ranges for which no residues were found and regular-style 2 user WUs were produced. The original double check thread has more info on the original situation.
Residues are built from steps in the calculation. The probability that two computers make the exact same error(s) in the same place(s) of a workunit is infinitesimal.*
*A case could be made that two identical CPUs with identical hardware bugs would do the wrong thing and match (see: Pentium FDIV bug), but it's been decades since an error of that magnitude has surfaced in retail hardware. The high diversity of modern CPUs minimizes the possibility that these bugs go undiscovered, as a CPU type with PG-significant errors would consistently fail validation across the domain. (In fact, PG recently saw something like this with AMD Navi GPUs, though it was really a software incompatibility issue, which happens, particularly with AMD).
____________
Eating more cheese on Thursdays. | |
|
Message boards :
Seventeen or Bust :
The SOB Double Check will end... aka The Way Too Early Prediction Thread |
