• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

DC Vault

alan2308 said:
Our lowest totals on BOINC based projects are Lattice, uFluids@Home, SZTAKI, NQueens@Home, and yoyo@Home.
Click to expand...
I have a bunch of uFluids and NQueens points pending, so there should be a boost there once they meet their quorum.

Also started client Community TSC (The Rothberg Institute For Childhood Diseases dedicated to finding a cure for children suffering from Tuberous Sclerosis Complex)

Homepage: http://www.childhooddiseases.org/
Team Profile : [H]ardOCP
 
Well guys - I'd stay away from Sudoku for a little while. I turned in a few good WUs last Saturday but now they have released some HUGE WUs that I have my doubts are actually working. I have one that is 39 hours in and has been sitting at 16% for a long time. Says its 129 hours to completion. I suspended all of them but one. I'll wait and see if that one actually finishes before I waste a bunch of time on the rest of them in my queue.:rolleyes:
 
Razor just received credit for another Lattice WU moving himself back into #1 [H]orde position with the project. :p
 
The good thing about the Sudoku units is that they have a quorum of 1, so you don't have to worry about losing points after all that time. I think I've got three WUs in the queue, and one at about 60 hours or so and around 50%, so the time to completion seems pretty accurate.

I don't think I'm going to get any points on my current giant SZTAKI WU because too many people errored out. The Lattice WUs are still cranking. I had to bring my new machine down for a few hours last night to put it into a case and straighten out the video card issues, but it should be back for good now. I won't be able to do any more Lattice WUs on this machine once the current ones finish, because the the video cards chew up too much RAM address space (with all 8 installed, I have 1GB of RAM for Windows -- I really should have thought ahead and got the x64 version).
 
I'm working on two Lattice WUs. One has been running for well over 24 hours and should finish today, and the other is just starting. I can probably set up Lattice on a different machine going forward, and keep the smaller low-RAM projects like PrimeGrid and ABC@Home on my multi-GPU system. Or maybe I'll just set up four more Wieferich@Home processes -- I need to see how much RAM those take. It seems like most projects run fine in only 1GB, and Lattice was the only one to give me "Waiting for memory" messages. Right now I have half the 9800GX2 cards turned off to give me 2.25GB to finish the WUs.
 
SazanEyes said:
The good thing about the Sudoku units is that they have a quorum of 1, so you don't have to worry about losing points after all that time. I think I've got three WUs in the queue, and one at about 60 hours or so and around 50%, so the time to completion seems pretty accurate.

I don't think I'm going to get any points on my current giant SZTAKI WU because too many people errored out. The Lattice WUs are still cranking. I had to bring my new machine down for a few hours last night to put it into a case and straighten out the video card issues, but it should be back for good now. I won't be able to do any more Lattice WUs on this machine once the current ones finish, because the the video cards chew up too much RAM address space (with all 8 installed, I have 1GB of RAM for Windows -- I really should have thought ahead and got the x64 version).
Click to expand...

Well its good to know that those Sudoku WUs eventually progress. I was beginning to wonder what was going on! :)

8 GPUs in one box! :eek: Very nice. What are you going to do with those? RC5-72? ;)
 
SazanEyes said:
I can probably set up Lattice on a different machine going forward
Click to expand...
Lattice should be releasing 38K MARXAN-based work units "any day now". These have much lower memory requirements. I will not be able to run these here as the client only supports Windows.
 
Tobit said:
Lattice should be releasing 38K MARXAN-based work units "any day now". These have much lower memory requirements. I will not be able to run these here as the client only supports Windows.
Click to expand...

As soon as they're out I'll be all over them. I'm set to accept everything but GARLI with a resource share of 500. Same for proteins@Home, so give me some work already. :D
 
We're finally on the board for proteins@Home. Capt. Napalm came over to our team with existing credit. I've already fired off an update request for our DC-Vault page. That leaves us only 5 "No Team" projects. :D
 
alan2308 said:
As soon as they're out I'll be all over them. I'm set to accept everything but GARLI with a resource share of 500.
Click to expand...
My one Windows BOINC machine has been crunching away at SHA-1 for awhile now. However, it can't handle GARLI projects at all. I'd love to be able to set it up so all MARXAN WUs could go to that machine and GARLI WUs to go to my Core i7. However, short of setting up a second account, I just don't see how this would be possible. Although, maybe I could do this through the various profiles. Off to take a look.

Edit: Yeah, this should work. For Lattice, I can set my Windows box to use my "work" profile and I have the "work" profile to only run MARXAN projects. :cool:
 
alan2308 said:
That leaves us only 5 "No Team" projects. :D
Click to expand...
We should have our PSP-Sieve team created soon, just waiting to hear back from the admin. He's currently traveling on business so may take another day or two. I've already finished one 100G Sieve range and working on my second. No factors found yet.
 
Razor_FX_II said:
I feel your pain I just completed a unit that had a quorum with 10 people that had client error.
That one was 642.05 claimed credit.:(
Click to expand...

Tobit said:
Ditto. Mine was 826.82 claimed and I'm going to get 0. I also have another one in progress that no one else was able to complete. I've turned in 7 units so far, only 1 has been granted credit to date. The other 5 still pending will hopefully be granted soon. I still have 7 WUs still crunching. I'm currently devoting 5 cores at a time to Lattice until I buy some more "D".
Click to expand...

I was going to post a link to this: http://boinc.umiacs.umd.edu/forum_thread.php?id=379 where the mod said that even those results would be used, then I did a double-take on the poster... Tobit, let us know if you get the credit for those WU that nobody else completed. Of the 4 I did, 3 had comp errors within minutes, and when I look at the WU detail, ever one had multiple other Comp errors, with no pending anywhere on the board...

Good news is that we're on the board in the Vault, with 3500 pernts in that project already, and we just hit #49!!! 7 places in two updates FTW! And at this rate we'll be at #47 after the next update...
 
Tobit said:
We should have our PSP-Sieve team created soon, just waiting to hear back from the admin. He's currently traveling on business so may take another day or two. I've already finished one 100G Sieve range and working on my second. No factors found yet.
Click to expand...

Good call, I already forgot about that. We're going to be pretty much down to malariacontrol.net here shortly.
 
Jathanis said:
I was going to post a link to this: http://boinc.umiacs.umd.edu/forum_thread.php?id=379 where the mod said that even those results would be used, then I did a double-take on the poster... Tobit, let us know if you get the credit for those WU that nobody else completed. Of the 4 I did, 3 had comp errors within minutes, and when I look at the WU detail, ever one had multiple other Comp errors, with no pending anywhere on the board...

Good news is that we're on the board in the Vault, with 3500 pernts in that project already, and we just hit #49!!! 7 places in two updates FTW! And at this rate we'll be at #47 after the next update...
Click to expand...


Good lord check out that Community TSC bump. We crossed over the line into the top 50 with 19 days to spare. Hold on, there's something in my eye. :D

47 isn't that unthinkable seeing as we'll have something adding on for proteins@Home hopefully on the next update.
 
Tobit said:
What a slacker, reason #24 I did not vote for him. :p
Click to expand...

Wow...... what where the other 23 reasons :(


I added the team and all my points followed me, so all better.

Thanks for catching that, This is all new to me over here.
 
Mr Flibble said:
Yeah 1000 unit's @ 1 point per unit but it still beats the hell out of running it on a CPU.
Click to expand...
My RC5-72 just got a new buffer of 15 that are worth 64 points each.
They take much longer now and the GPU heats up more since it runs continuous now.
 
Awesome, we will be on SoB-Sieve and PSP-Sieve real soon. I just found one factor 75% through my second 100G range which should complete in ~2 hours.

Code: 
jt@ganymede:~/sr2sieve$ ./sr2sieve-x86_64-linux -s -v
sr2sieve 1.8.6 -- A sieve for multiple sequences k*b^n+/-1 or b^n+/-k.
Compiled on Jan 16 2009 with GCC 3.4.6 (Debian 3.4.6-1).
SSE2 code path, L1 data cache 32Kb (detected), L2 cache 256Kb (detected).
Read 1827430 terms for 15 sequences from dat format file `SoB.dat'.
Sieving standard form sequences k*2^n+/-1.
Split 15 base 2 sequences into 140 base 2^360 subsequences.
Using 2373Kb for subsequence bitmaps.
Using 310Kb for subsequence congruence tables.
Using 446Kb for Legendre symbol lookup tables.            
Range end is too high for SSE2 code path, switching to x87 FPU.
Using 32Kb for the baby-steps giant-steps hashtable, maximum density 0.27.
Baby step method gen/8, giant step method new/4, ladder method gen/8.
Using 128Kb for the Sieve of Eratosthenes bitmap.
Expecting to find factors for about 0.39 terms in this range.
sr2sieve 1.8.6 started: 991 <= n <= 49999997, 12681350000000000 <= p <= 12681450000000000
12681424752549583 | 24737*2^3827023+1                                      
p=12681425105304629, 2946876 p/sec, 1 factor, 75.1% done, ETA 26 Jan 23:09
 
Kendrak said:
Thanks for catching that, This is all new to me over here.
Click to expand...

Its pretty much all new to us too. We're still making a lot of it up as we go along. :D

Which remind me, I need to do some more ABC, I hate being last. Especially when last is 7th now. :rolleyes:
 
Tobit said:
Awesome, we will be on SoB-Sieve and PSP-Sieve real soon. I just found one factor 75% through my second 100G range which should complete in ~2 hours.
Click to expand...
If you get the chance could you give some instructions and links on how to get these projects running?
 
Razor_FX_II said:
If you get the chance could you give some instructions and links on how to get these projects running?
Click to expand...
Alan summed it up well on page 11 of this thread, I stumbled through it but some things could be clarified a little more as there are many steps involved. Let me write something up, give me a little bit. :cool:
 
Tobit said:
Alan summed it up well on page 11 of this thread, I stumbled through it but some things could be clarified a little more as there are many steps involved. Let me write something up, give me a little bit. :cool:
Click to expand...
Thanks, I appreciate it, as I would like to get up and running on these as well but got confused when I tried to figure it out last time.
 
Razor_FX_II said:
My RC5-72 just got a new buffer of 15 that are worth 64 points each.
They take much longer now and the GPU heats up more since it runs continuous now.
Click to expand...

Yeah i got some of these too. Good to see the GPU running constant though instead of so many peaks and valleys.

Nice job on breaking into the Top 50 guys! Next stop - Top 40! And its probably not all that far off. :)
 
Ok folks, to get going on SoB-Sieve and PSP-Sieve, here are the steps involved:

  • Create an account on the reservation site.
  • Log in and reserve a range.
  • Install the latest sr2sieve software. This can be found here.
  • Download the DAT file to the sieve folder.
  • In your sieve folder, create a text file 'sr2work.txt' with the range assigned to you in the following format:
    Code: 
    jt@ganymede:~/sr2sieve$ more sr2work.txt 
12681350-12681450
  • Launch sr2sieve with the "-s" flag. Other options are available, read the READMEs.
  • Once the range is complete you mark it as complete on the reservation site and submit your data to the Seventeen or Bust site AND the Prime Sierpinski Project. Make sure you are logged into the SoB site before entering your factors. Additionally, for PSP, you need to make a request to have your name added to the import page before you can submit for the first time. You can do this in this thread here.
Sieving is not for the faint of heart as this is a very manual project. When you reserve a range please ensure that you will see it through to completion and not abandon it. The maintainers ask that you please update you reserve ranges ever 30 days with a comment to let them know you are still alive.

Edit: When a factor is found, a text file is created called factors(range start no.).txt <- this is the data you need to submit to the two sites. If a factor is not found, you do not submit anything. You simply mark the range as completed on the reservation site.

Edit2: Sample of a factors result file:
Code: 
jt@ganymede:~/sr2sieve$ more factors12681350.txt 
12681424752549583 | 24737*2^3827023+1
 
If plans stick, I should have a second i7 busy with WCG in the coming month or two.
 
Razor_FX_II said:
My RC5-72 just got a new buffer of 15 that are worth 64 points each.
They take much longer now and the GPU heats up more since it runs continuous now.
Click to expand...


good to see they have the gpu client back.. cant wait til the first of the month so i can start crunching again.. arg the wait is killin me! but so far already did the math on the meter and the electricity bill should be about 250 dollars this month instead of 400 dollars like last month.. which means im probably going to retire my electricity hogging 3100+ since full load with the overclock uses over 150w just to the cpu.. not counting everything else..
 
Mr Flibble said:
I guess I can stop complaining about my $122 PG&E bill.
Click to expand...


122 dollars.. where the hell do you live? i want a 122 dollar PG&E bill.. hell mine was 400 dollars not counting the 150 dollars for gas added to that..
 
Jathanis said:
Really? what city? I'm over in the Stockton/Modesto area... Drive your way all the time!
Click to expand...

im in oakley.. which is just outside of the bay area.. even though its considered part of the east bay.. i lived in stockton for 16 years.. right off hammer lane near I-5
 
I lived in Dublin for 8 years (went to High School out there years back)

I'm in Ohio and out bill is coming out to ~150 a month now. I'm a bit worried about summer. :eek:
 
Up 2 more spots to #47. Congrats gentlemen, and I'm still not sick of this post almost every morning! :D
 
My points to uFluids still haven't posted (they are all pending) but when they do that should help.

The one core is still working away at the same climate WU. A whole 10% !!!!

It has posted points though, so I am not going to complain.

NQueens@Home does not seem to give a bunch of points, but I'm chipping away at that as well as ABC.

ABC rocks the points!
 
Tobit said:
In your sieve folder, create a text file 'sr2work.txt' with the range assigned to you in the following format:
Code: 
jt@ganymede:~/sr2sieve$ more sr2work.txt 
12681350-12681450
Click to expand...
I used:
Code: 
name@provider.com:~/sr2sieve$ more sr2work.txt 
12682750-12682850
but got an error in the log file that it wasn't properly formatted so I used:
Code: 
12682750-12682850
And its up and running now. Hopefully this is correct.

Thanks for your instructions that really helped.:cool:
 
You must log in or register to reply here.
Back
Top