mersenneforum.org

mersenneforum.org (https://www.mersenneforum.org/index.php)
-   No Prime Left Behind (https://www.mersenneforum.org/forumdisplay.php?f=82)
-   -   Sieving drive for k=1005-2000 n=200K-500K (https://www.mersenneforum.org/showthread.php?t=11204)

gd_barnes 2008-12-26 09:10

Sieving drive for k=1005-2000 n=200K-500K
 
[COLOR=black][FONT=Verdana]This is a sieving drive for k=[/FONT][/COLOR][COLOR=black][FONT=Verdana]1005-2000 and n=200K-500K. If you can spare the machines, we are asking that people dedicate 15-20% of their CPU resources to the effort so that we can complete it in ~3 weeks. I will personally be putting 2 of my newer 64-bit Intel quads on it.[/FONT][/COLOR]

The reason for the priority is because the 5000th-place prime on the top-5000 site is rising rapidly, largely as a result of us finding so many primes on our 1st drive and now the 5th/6th/7th drives. :smile: We would like to begin LLRing the effort before the 5000th place prime reaches n=350K, which at the current rate will be in about 3 weeks.

Bruce started the sieve on n=50K-500K several months ago and got it up to P=160G. Ian has accelerted the pace in the last 2 months. Our current sieve depth is P=1400G and Ian is currently working on P=1400G-1800G. I have determined the optimal sieve depth for the n=50K-200K range to be about P=1600G so we will break off that range and not include it in this effort. In the mean time, we need to continue sieving n=200K-500K starting from P=1800G.

As you can guess, even for the smaller range of n=200K-500K, the file is huge. Even in ABCD format, the entire n=50K-500K range is 32 MB and the n=200K-500K range that we will be working on is 21 MB after removing all factors up to P=1400G. Zipped, it should be about 6 MB. I provide a link to the zipped file below. If it is a problem for you to pull it off of the link, let me know and I'll Email it to you.

We will likely sieve to about P=6T (6000G) and then break off n=200K-350K. We'll then need to continue sieving n=350K-500K. A rough estimate for the optimum sieve depth for that final range is P=9T-10T (9000G-10000G).

Ian has been sieving P=200G per core and it has been taking him ~10-11 days using 64-bit sieving. Assuming that you have new cores and a 64-bit O.S., I would suggest taking P=100G ranges per core to start with to guage the sieving speed of your machine. If his machines are a guage, that would take you 5-6 days for each core. For slower/older machines or if you. have a 32-bit O.S., I would suggest half of that.

sr2sieve is what we will use. Let us know if you need the executable or more detailed instructions on using it.

I prefer that we use the more "modern" method of sieving by specifying the ranges and file name at the command prompt as in:

sr2sieve -p 1800e9 -P 2000e9 -i sieve1005-2000-200K-500K.txt

The above would be if you were sieving P=1800G-2000G. The file is listed after the "-i" command and is the actual file name that will be in the zipped file at googlepages. Feel free to name it something shorter if you want or use the "srwork" older convention where you don't have to specify a file name.

When complete, you should have a factors.txt file. Just post the file here in this thread or if it is too big, please Email the file to me at:
gbarnes017 at gmail dot com

Sieve file:
[URL="http://gbarnes017.googlepages.com/sieve1005-2000-200K-500K.zip"]1005-2000-200K-500K[/URL] [factors removed up to P=3800G]


Reservations:
[code]
P-range reserved by Status
n=50K-500K:
0G- 160G Brucifer complete
160G-1800G MyDogBuster complete

n=200K-500K:
1800G-3400G gd_barnes complete
3400G-3800G Flatlander complete
3800G-4000G MyDogBuster complete
4000G-4050G Mini-Geek complete
4050G-5850G MyDogBuster complete
5850G-5900G mdettweiler complete
5900G-7100G gd_barnes complete
7100G-7200G Mini-Geek complete
7200G-8200G IronBits complete
8200G-8600G gd_barnes complete
8600G-8800G glennpat complete
8800G-10000G gd_barnes complete
[B]It won't be neccesary to reserve ranges for P>10000G.[/B]
[/code]After Ian is complete to P=1800G, we will start an 8th drive for n=50K-350K beginning with the range of n=50K-200K. After we have completed all sieving, we will start a 9th drive for n=350K-500K. n=200K-350K will be done after n=50K-200K in the 8th drive.

There's a lot of work to do so let's get crackin'. Once this is done, we'll have a drive where we'll likely be finding a gob of top-5000 primes each day! :smile:


Gary

gd_barnes 2008-12-26 09:13

Reserving P=1800G-3400G to put on 8 cores. The sieving speed on my Intel machines is a bit faster than the speed of Ian's sieve due to removing the n=50K-200K range and it being a higher P-range. Sr2sieve gives the ETA for the P=1800G-2000G range on one core as Jan. 5th so I'm loading all 8 of 'em up for 10 days! :-)

2 quads are coming off the 5th drive, which has been good to me so far. Boo-hoo! lol For the first time in 2+ months, I'll have no cores on port 400. 4+ are still on port 5000 though.


Gary

Flatlander 2008-12-26 15:32

Taking 3400-3600G.

Another one for the book:
"Never knowingly undersieved." :smile:

I'm running the above on two cores, 32 bit Windows, 3.15GHz. ETA 1st Jan.


This would be an ideal opportunity for me to try out the 64bit Ubuntu I have installed on my C2Duo.
I don't really want to start learning to speak fluent Linux just yet, but perhaps someone could post instructions here for me to get two instances of sr2sieve running? (Or maybe send me a PM.)
The instructions would need to be [I]very [/I]specific. :smile: It would defeat the object if the PC was idle for hours while I figure things out.

nuggetprime 2008-12-26 16:30

Why are we not using srsieve? For that lot of k's it should be faster.

--nugget

mdettweiler 2008-12-26 17:41

[quote=Flatlander;155140]Taking 3400-3600G.

Another one for the book:
"Never knowingly undersieved." :smile:

I'm running the above on two cores, 32 bit Windows, 3.15GHz. ETA 1st Jan.


This would be an ideal opportunity for me to try out the 64bit Ubuntu I have installed on my C2Duo.
I don't really want to start learning to speak fluent Linux just yet, but perhaps someone could post instructions here for me to get two instances of sr2sieve running? (Or maybe send me a PM.)
The instructions would need to be [I]very [/I]specific. :smile: It would defeat the object if the PC was idle for hours while I figure things out.[/quote]
Well, sr2sieve under Linux, both for 32-bit and 64-bit, is in fact quite identical in almost every way to the Windows version. First, you need to download the latest 64-bit Linux version of sr2sieve, [url=http://www.geocities.com/g_w_reynolds/sr2sieve/sr2sieve-1.7.15-linux-x86-64.tar.gz]here[/url]. Make a directory somewhere in a convenient place for sr2sieve, and extract the file you just downloaded there. Then, download the sieve file from this thread and put it in that folder too.

Now open a terminal window and navigate to your sr2sieve folder using the "cd" command. This is very similar to cd/chdir in DOS; the only difference is that capitalization DOES matter, and that you should use forward slashes instead of backslashes. For example, if sr2sieve is in the directory "sr2sieve" on your desktop, type the following command:
[I]cd /home/username/Desktop/sr2sieve[/I]
Replavce "username" with the username that you use to log in (that is, the lowercase, short one, not the full name that Ubuntu also has you enter when setting up).

Once you've gotten to your sr2sieve directory, you can run sr2sieve pretty much the same way you do under Windows. Th eonly difference is that you have to preface the commands with "./" (without quotes). For example, if under Windows you would have executed the following command:
[I]sr2sieve -p 1800e9 -P 2000e9 -i sieve1005-2000-200K-500K.txt[/I]
under Linux you would run the following instead:
[I]./sr2sieve -p 1800e9 -P 2000e9 -i sieve1005-2000-200K-500K.txt[/I]

If you'd like to run multiple instances, simply do it whichever way you were doing it under Windows. Both the -u method and running from separate folders will work just as well under either operating system. :smile:

Hope this helps! Feel free to drop me a PM if you've got any questions. :smile:

Max :smile:

mdettweiler 2008-12-26 17:42

[quote=nuggetprime;155145]Why are we not using srsieve? For that lot of k's it should be faster.

--nugget[/quote]
I think srsieve *used* to be faster for this many k's, but modifications introduced in recent versions of sr2sieve have in fact made sr2sieve faster. Feel free to try it for yourself and compare the p/sec. time to see which one's faster on your machine.

gd_barnes 2008-12-26 19:59

[quote=nuggetprime;155145]Why are we not using srsieve? For that lot of k's it should be faster.

--nugget[/quote]

No. With the newer versions of sr(x)sieve, sr2sieve is definitely faster.

[quote=mdettweiler;155148]I think srsieve *used* to be faster for this many k's, but modifications introduced in recent versions of sr2sieve have in fact made sr2sieve faster. Feel free to try it for yourself and compare the p/sec. time to see which one's faster on your machine.[/quote]

Yes, nugget, feel free to give both a try if you want. I think the latest version of srsieve is 0.6.10 and for sr2sieve is 1.7.15. Be sure and use these latest 64-bit versions when comparing.


Gary

gd_barnes 2008-12-26 20:04

[quote=Flatlander;155140]Taking 3400-3600G.

Another one for the book:
"Never knowingly undersieved." :smile:

I'm running the above on two cores, 32 bit Windows, 3.15GHz. ETA 1st Jan.


This would be an ideal opportunity for me to try out the 64bit Ubuntu I have installed on my C2Duo.
I don't really want to start learning to speak fluent Linux just yet, but perhaps someone could post instructions here for me to get two instances of sr2sieve running? (Or maybe send me a PM.)
The instructions would need to be [I]very [/I]specific. :smile: It would defeat the object if the PC was idle for hours while I figure things out.[/quote]


I'm still pretty much a Linux newb and using Max's instructions several months back was easily able to get 64-bit sr(x)sieve running on my Linux machines.

You just have to have Ubuntu version 8.04 or later installed on your machines to get the increased speed of 64-bit processing. Correct me if I'm wrong there Max.

Come on Max, Carlos, and co., let's get some sieving reservations going here. Ian, can you add another quad to the mix also? :smile: It's OK if some of your reservations here are delayed by a couple of weeks.


Gary

Flatlander 2008-12-26 20:30

Taking/running 3600-3800G on two cores in Ubuntu 64bit. :smile:


I am getting little/no speed increase over 32bit Windows. Where can I get details of CPU speed/temps etc? How do I assign affinity to each core?

em99010pepe 2008-12-26 20:43

How many k's are we talking about? How many are we double checking?

Carlos

MyDogBuster 2008-12-26 20:46

I'm taking 3800 to 4200G

mdettweiler 2008-12-26 20:46

[quote=gd_barnes;155166]I'm still pretty much a Linux newb and using Max's instructions several months back was easily able to get 64-bit sr(x)sieve running on my Linux machines.

You just have to have Ubuntu version 8.04 or later installed on your machines to get the increased speed of 64-bit processing. Correct me if I'm wrong there Max.[/quote]
One correction: you don't necessarily have to have version 8.04 or later. The only requirement is that you have a 64-bit version--and 64-bit versions of Ubuntu have been available long before 8.04. It's just that at the time you ordered your CD to install 64-bit on your machines with, 8.04 was the latest version so that's what you got. :smile: (There has been a new version, 8.10, released since then, but 8.04 should be good for a while since it's a "Long Term Support" version, meaning that Ubuntu will keep providing updates on it for a longer duration than usual. Thus, the Long Term Support versions such as 8.04 are usually good choices for machines like yours where the priority is stability and longevity, not necessarily cutting-edge versions.)
[quote]Come on Max, Carlos, and co., let's get some sieving reservations going here. Ian, can you add another quad to the mix also? :smile: It's OK if some of your reservations here are delayed by a couple of weeks.[/quote]I should be finishing my 5th Drive reservation sometime within the next day or so. At that point I'll grab a sieving range to replace it with. :smile:

Also, some good news: my quad is finally being put into regular use today! :grin: I'm transferring all the files from the old computer that it's replacing, and then it should be on for an average of at least 12-14 hours a day, hopefully more. :smile: Unfortunately, though, since it's not always the most convenient to manage prime search programs on it, it's primarily limited to LLRnet (i.e. no manual LLR or sieving). I've got it on G4000 for the time being.

Max :smile:

MyDogBuster 2008-12-26 21:25

I have to shuffle things around after I return from an appointment. I'll be reserving more then. Let's try to get it to 8T by Jan 24th.

Mini-Geek 2008-12-26 21:55

I'll take 4000G-4050G. This will delay my Drive 5 reservation a little (until it's done, obviously).

MyDogBuster 2008-12-26 22:33

The reservations in the first post are wrong. As Flatlander pointed out to me, he reserved 3600-3800. I changed mine to 3800-4200.
Just trying to keep the books straight.

em99010pepe 2008-12-26 22:36

Ian, change yours to 3800-4000 and then pick up from 4050.

MyDogBuster 2008-12-26 23:12

[quote]Ian, change yours to 3800-4000 and then pick up from 4050.
[/quote]

Done

It looks like this (I hope)

Flatlander 3400-3800
MyDog 3800-4000
MiniGeek 4000-4050

gd_barnes 2008-12-26 23:23

[quote=Flatlander;155140]Taking 3400-3600G.

Another one for the book:
"Never knowingly undersieved." :smile:

I'm running the above on two cores, 32 bit Windows, 3.15GHz. ETA 1st Jan.


This would be an ideal opportunity for me to try out the 64bit Ubuntu I have installed on my C2Duo.
I don't really want to start learning to speak fluent Linux just yet, but perhaps someone could post instructions here for me to get two instances of sr2sieve running? (Or maybe send me a PM.)
The instructions would need to be [I]very [/I]specific. :smile: It would defeat the object if the PC was idle for hours while I figure things out.[/quote]

[quote=Flatlander;155174]Taking/running 3600-3800G on two cores in Ubuntu 64bit. :smile:


I am getting little/no speed increase over 32bit Windows. Where can I get details of CPU speed/temps etc? How do I assign affinity to each core?[/quote]

[quote=MyDogBuster;155179]I'm taking 3800 to 4200G[/quote]

[quote=MyDogBuster;155187]The reservations in the first post are wrong. As Flatlander pointed out to me, he reserved 3600-3800. I changed mine to 3800-4200.
Just trying to keep the books straight.[/quote]



What the heck?

Chris, you reserved 3400-3600 and then 3600-3800.

Ian, I posted that Chris originally reserved 3400-3600 and hadn't checked this thread since then. How could they have been wrong? lol

Ian, you reserved 3800-4200 and then said you're changing to 3800-4200. lmao

Have people gone loony here?

Please check post 1 here and let me know if it is right now.


Gary

gd_barnes 2008-12-26 23:26

OK, I can kind of see what happened now. It looks like Ian tried to reserve 3600-4000 and Max changed the 1st post here after Chris had already reserved 3600-3800 and THEN Ian changed his original post to 3800-4200 making me wonder how he could change 3800-4200 to 3800-4200. lol Too many cooks in the kitchen.

The only question I have left is, Ian, are you taking 4050-4200 or 4050-4250 since Mini-Geek is doing 4000-4050?

mdettweiler 2008-12-26 23:27

[quote=MyDogBuster;155192]Done

It looks like this (I hope)

Flatlander 3400-3800
MyDog 3800-4000
MiniGeek 4000-4050[/quote]
Okay, fixed. :smile:

P.S.: It looks like Gary is on at the same time as I and posted his message about this as I was fixing the first post of this thread. Gary, just wanted to let you know so that we don't step all over each other's edits. :smile:

gd_barnes 2008-12-26 23:36

[quote=em99010pepe;155177]How many k's are we talking about? How many are we double checking?

Carlos[/quote]

All of them, 498 total. For the n=50K-350K range, < 10% double-check. For the n=350K-500K range, about 30% double-check. The double-checking of the top-5000 range while more than the 1st drive is much less than the 3rd drive where we had slightly > 50% double-check for n>480K and 35-40% double-check for n=260K-480K.

[quote=Mini-Geek;155186]I'll take 4000G-4050G. This will delay my Drive 5 reservation a little (until it's done, obviously).[/quote]


Not a problem. That's to be expected. Everyone's reservations that help out here will be delayed a little.


Gary

gd_barnes 2008-12-26 23:40

[quote=MyDogBuster;155183]I have to shuffle things around after I return from an appointment. I'll be reserving more then. Let's try to get it to 8T by Jan 24th.[/quote]


Let's shoot for P=9T by Jan. 15th-17th. Once my P=1600G range is done by Jan. 5th, I'll be able to take another P=1600G range and finish it by Jan. 15th. Ian, if you can do the same, with others help, we should be able to get there. If not, I'll think about pulling 2 more quads in on the effort although would prefer not to if possible.

I think the 5000th prime will be at n=350K right around Jan. 15th-20th.


Gary

mdettweiler 2008-12-26 23:40

[quote=Flatlander;155174]Taking/running 3600-3800G on two cores in Ubuntu 64bit. :smile:


I am getting little/no speed increase over 32bit Windows. Where can I get details of CPU speed/temps etc? How do I assign affinity to each core?[/quote]
Hmm...that's weird. Are you sure you downloaded the sr2sieve version labeled "x86_64" for Linux instead of the one labeled just "x86"? Otherwise you may be in fact running the 32-bit Linux version under 64-bit Ubuntu--which would explain why you're not seeing any speed increase. There should be about a 2x increase in the p/sec. rate for 64-bit vs. 32-bit.

As for affinity: try running the command ./sr2sieve -h in your sr2sieve directory. That's the same as "sr2sieve -h" on Windows; it will give you a complete listing of all the command line flags that sr2sieve supports, including ones that set affinity.

MyDogBuster 2008-12-27 00:01

Okay, the scoreboard seems to be correct SO,

I'll take 4250 - 5850. That should fill me up till about the Jan 16th.

gd_barnes 2008-12-27 00:25

[quote=MyDogBuster;155216]Okay, the scoreboard seems to be correct SO,

I'll take 4250 - 5850. That should fill me up till about the Jan 16th.[/quote]


I now have you down for the entire range of 4050 - 5850.

Flatlander 2008-12-27 01:26

[quote=mdettweiler;155209]Hmm...that's weird. Are you sure you downloaded the sr2sieve version labeled "x86_64" for Linux instead of the one labeled just "x86"? ....[/quote]
It was the 64bit version from the link you provided. I assume if I've accidentally installed 32bit Ubuntu, 64 bit sr2sieve wouldn't run at all?

MyDogBuster 2008-12-27 01:31

[QUOTE]I now have you down for the entire range of 4050 - 5850.
[/QUOTE]

That's right. It's a mess. I can still get about 400G for my original 2 cores, but I'll hold off on that to see how everything goes.

mdettweiler 2008-12-27 01:49

[quote=Flatlander;155222]It was the 64bit version from the link you provided. I assume if I've accidentally installed 32bit Ubuntu, 64 bit sr2sieve wouldn't run at all?[/quote]
You are correct--32-bit Ubuntu would not run 64-bit sr2sieve at all. And I just verified that the link I posted *is* for the latest 64-bit version of sr2sieve--so I have absolutely no idea why it's acting this way for you.

I would suggest posting about your situation in [url=http://www.mersenneforum.org/showthread.php?t=5785]this thread[/url], which is where Geoff usually posts general information about the sr*sieve programs. Maybe he'll know what's going on here.

Flatlander 2008-12-27 03:30

[quote=mdettweiler;155226]You are correct--32-bit Ubuntu would not run 64-bit sr2sieve at all. And I just verified that the link I posted *is* for the latest 64-bit version of sr2sieve--so I have absolutely no idea why it's acting this way for you.

I would suggest posting about your situation in [URL="http://www.mersenneforum.org/showthread.php?t=5785"]this thread[/URL], which is where Geoff usually posts general information about the sr*sieve programs. Maybe he'll know what's going on here.[/quote]
Is there some way I can get a reading of the speed I'm running at. (In Ubuntu not in the BIOS.) I've downloaded "Device Manager" but it just tells me the stock speed.

mdettweiler 2008-12-27 06:31

[quote=Flatlander;155239]Is there some way I can get a reading of the speed I'm running at. (In Ubuntu not in the BIOS.) I've downloaded "Device Manager" but it just tells me the stock speed.[/quote]
Okay, right-click on one of your taskbars (the upper taskbar is better for this since adding gadgets there won't hog room from your active-window buttons), and click Add to Panel. Locate the CPU Frequency Scaling Monitor in the list that comes up, and click Add to Panel twice (once for each core). Then right-click on the second gadget that you just added and click Properties; set it to report the frequency for CPU 1 (the other should be left at the default, CPU 0). That should tell you what speed your CPU is currently running at.

If it's running lower than it should, you can use the CPU Frequency Scaling Monitors to fix this. Open a terminal window, and run the command "sudo dpkg-reconfigure gnome-applets" (without quotes). Enter your password when prompted. It will ask you whether or not to allow the cpufreq-selector applet to have root privileges; select Yes and press Enter. You will be dumped back at the terminal prompt; now close the terminal.

Now, if you left-click on the applet, you'll see options to manually set the CPU to a particular frequency, as well as multiple predefined settings (referred to as "governors")--namely, "Conservative", "Ondemand", "Performance", and "Powersave". Choose either the highest clock speed level, or the "Performance" governor, and you should be in business. :smile:

Hope this helps! :smile:

Max :smile:

Flatlander 2008-12-27 07:13

Thanks Max. :smile:
That showed me that it was running at stock (1.8GHz) instead of 2.38GHz. For some reason the BIOS had reset the speed when I rebooted.
I'm now getting slightly better p/sec than my quad @ 3.15Ghz.

So 64bit is running about 33% faster than 32bit. (Maybe the caches are different. I'll check it out later.)

mdettweiler 2008-12-27 07:28

[quote=Flatlander;155261]Thanks Max. :smile:
That showed me that it was running at stock (1.8GHz) instead of 2.38GHz. For some reason the BIOS had reset the speed when I rebooted.
I'm now getting slightly better p/sec than my quad @ 3.15Ghz.

So 64bit is running about 33% faster than 32bit. (Maybe the caches are different. I'll check it out later.)[/quote]
Okay, glad to hear you got it working. :smile: Though, I can't imagine how the caches would be different between operating systems...I would instead presume that the boost only being 33% is due entirely to the 64-bit system being clocked at 2.38Ghz rather than the 3.15Ghz 32-bit comparison machine.

mdettweiler 2008-12-27 08:16

Taking 5850G-5900G. I'll start this tomorrow (my 5th Drive range should finish overnight). :smile:

Mini-Geek 2008-12-30 12:49

1 Attachment(s)
4000G-4050G complete, factors attached. I'll sieve another 50G after I finish the next section in my main drive processing, which should be on the 2nd. For now I'm not reserving any more.

gd_barnes 2008-12-30 13:39

With the primes just pouring in at top-5000 as a result of PrimeGrid running an effort on the Proth side similar to what we are currently for k<1200 and at lower n-ranges, it looks like n=350K will be at n=350K by Jan. 5th or earlier.

Since people reserved and divided up their ranges on their cores assuming a Jan. 15th completion date, we'll stick with our plans here but I'll put another quad on it starting later today or on Weds that will boost the reserved ranges to P=~7T. On Jan. 5th, my 1st 2 quads will finish and I'll be able to do another P=1.6T range by Jan. 15th. That would put it at P=8.6T so we'd likely just need a few more smaller reservations to get us where we need to be.

The drive will still be n=350K-500K but the first few primes may be non-top-5000. I'm now estimating that the 5000th prime will be at n=351K by the time we start.

As if we hadn't done so already, PrimeGrid has demonstrated conclusively that our method of searching for primes is the wave of the future! :smile:


Gary

MyDogBuster 2008-12-30 15:29

[quote]Ian, how far from completion are you on the P=1.4T-1.8T range? If you're < 3 days from completion, we'll wait on it. If we're gonna be undersieved for part of it, we may as well minimize how much we are undersieved by. lol[/quote]

My trusty countdown timer says that my 1.4-1.8T sieves won't be done until Jan 4th about 10PM, about 5 1/2 days.

I'm up to 1.506T on the 1.4-1.6 sieve.

Flatlander 2008-12-30 15:30

[quote=Flatlander;155800]...
My reservation will finish late 01/01/09 GMT. (Looks like >20,000 factors.)[/quote]
In case you didn't see my second edit.

gd_barnes 2009-01-01 04:29

I moved all of the posts specifically related to the now-running n=350K-352K range to the 8th drive thread.

gd_barnes 2009-01-01 10:03

Reserving P=5900G-7100G. I'm adding a quad to the effort. ETA is early on Jan. 16th.

My current range on 2 quads will be done very early on Jan. 6th. At that point, I think we'll have everyone's factors up to either P=3800G or P=4050G. We'll remove all the factors from the file and calculate an optimum sieve depth for the entire range. I'm estimating it will be P=9000G-10000G somewhere.

With the above 2 quads done on Jan. 6th, I'll be able reserve another P=1600G range that should also complete on Jan. 16th. Taking that into account, we'd likely only need another P=500G-1000G worth of reservations from everyone to complete the sieving.

With the pressure off now, Jan. 20th is still what we're shooting for but if we can be done a little early, it'll be that much sooner we can get down to some serious prime finding. :smile:


Gary

Flatlander 2009-01-01 15:59

1 Attachment(s)
3400-3600G complete.
3600-3800G will finish in a few hours.

Flatlander 2009-01-02 13:47

1 Attachment(s)
3600-3800G complete.

MyDogBuster 2009-01-03 03:32

We could use more help in the sieving effort. The quicker we get the sieving done, the quicker we get pairs loaded to test. I, for one, hate wasting time testing pairs that I shouldn't have to.

We are booked to 7.1T, we need to get to 9T to have an optimal sieve.

Everyone testing the pairs should help with the sieve.

Mini-Geek 2009-01-03 03:58

Reserving 7100G-7200G.

IronBits 2009-01-03 04:45

I'd love to help, but I don't recall how to do a darn thing with sieve, nor what software or how to go about it.
I asked Max to send me a care package that I can unzip to some directory and run some program on 8 cores, Intel I7, Vista 64bit.
Let's see what happens...

MyDogBuster 2009-01-03 04:54

[quote]I'd love to help, but I don't recall how to do a darn thing with sieve, nor what software or how to go about it.
I asked Max to send me a care package that I can unzip to some directory and run some program on 8 cores, Intel I7, Vista 64bit.
Let's see what happens
[/quote]

Care packages are good. LOL

BTW I also got an Intel I7 last Monday. I finally had enuf of my AMD Quad.
It has 4 cores, but 8 threads. I figured out that it is the equivalent of a 6 core quad running at 2.66ghz each. SWEET

IronBits 2009-01-03 05:06

:party:
That's excellent!
Now if I could only convince you to join Free-DC, if for nothing more than some balance. :grin:

mdettweiler 2009-01-03 06:26

[quote=IronBits;156587]I'd love to help, but I don't recall how to do a darn thing with sieve, nor what software or how to go about it.
I asked Max to send me a care package that I can unzip to some directory and run some program on 8 cores, Intel I7, Vista 64bit.
Let's see what happens...[/quote]
Okay, given that...I'll reserve 7200G-8200G for you and send you a package shortly all pre-configured for that range. That should hold you for about a week on all 8 cores, if my figuring is correct. :smile:

IronBits 2009-01-03 06:43

Hey Dog, I'm using an
ASUS P6T Deluxe mobo
Monsoon III LT HSF
6gb Corsair TR3X6G1333C9 R memory (3x2gb)
Just got mine running on Vista 64, what are you using?
Have you pushed yours to 3.8GHz yet?
What's yer rig dude? Spill the beans :grin:

Thinking about dual booting with 2nd HDD into Linux ;)

IronBits 2009-01-03 07:19

Thanks Max!
Sieving has commenced. :)

MyDogBuster 2009-01-03 07:24

[QUOTE]Hey Dog, I'm using an
ASUS P6T Deluxe mobo
Monsoon III LT HSF
6gb Corsair TR3X6G1333C9 R memory (3x2gb)
Just got mine running on Vista 64, what are you using?
Have you pushed yours to 3.8GHz yet?
What's yer rig dude? Spill the beans :grin:

[/QUOTE]

I don't overclock.

I only have 3gb DDR3 memory 1066 mhz

Not a clue on the mobo. Whatever was in there.

Vista 64

IronBits 2009-01-03 08:39

Sounds like a store bought rig. Probably simpler that way in many ways.
I bought a whole computer once, my first one, back in 1984-1985, 286-12 MHz with turbo button to toggle between 4.77 and 12, which was important when playing centipede. <lol>

nuggetprime 2009-01-03 13:16

Taking 8200G-8600G.

nugget

gd_barnes 2009-01-03 13:56

Thanks everyone for quickly hopping in here after the we got the n=350K-352K range ran!

To clarify so that everyone is clear:

1. We shouldn't need to sieve past P=10000G (10T). I just don't want anyone to waste any cores by reserving anything past that limit. I'll make a note in the 1st post of the thread about it.

2. My first range will be done early on Jan. 6th on 2 quads. I could then load them with another P=1600G range that would complete by Jan. 16th but I see that there is now only a P=1400G range remaining up to P=10000G (10T). Therefore on Jan. 6th, I'll reserve whatever is remaining to P=10T. At that point, the reservations will be closed for the sieving effort.

3. I am planning on all of my ranges being done on Jan. 16th. If you can shoot for that date on your reserved ranges, that would be helpful.


Technically, we're shooting for Jan. 20th but with all the help coming in, I figured we can shoot for sooner.


Thanks,
Gary

nuggetprime 2009-01-03 16:24

Just a suggestion: run multiple threads instead of multiple instances. It's easier to handle and is even a little bit faster because less RAM is used.

nugget

IronBits 2009-01-03 16:36

How do you do that?
I assume your talking about having several startup scripts, just one startup script?

nuggetprime 2009-01-03 16:43

No,sr2sieve has a -t option to run multiple threads. With that option you can just use one instance/folder/command and still use the 4 cores of a quad. Normally this is slower than running 4 instances,but when the sievefile is big,it can be faster.

mdettweiler 2009-01-03 17:17

[quote=nuggetprime;156662]No,sr2sieve has a -t option to run multiple threads. With that option you can just use one instance/folder/command and still use the 4 cores of a quad. Normally this is slower than running 4 instances,but when the sievefile is big,it can be faster.[/quote]
Really? Hmm...I didn't know multithreading could be faster with a big sieve file.

The "care package" that I put together to get David started yesterday runs multiple instances using the -u switch--there are eight batch files in the sr2sieve folder, each with 1/8 of the total range.

MyDogBuster 2009-01-03 18:08

[QUOTE]3. I am planning on all of my ranges being done on Jan. 16th. If you can shoot for that date on your reserved ranges, that would be helpful.
[/QUOTE]

It'll be tight for me but I think I'll sneak in just at the wire (assuming no glitches). As I say that, one of my quads doing the sieving just re-booted all by itself.

glennpat 2009-01-04 14:36

I'll take 8600-8800G.

gd_barnes 2009-01-04 15:23

[quote=MyDogBuster;156674]It'll be tight for me but I think I'll sneak in just at the wire (assuming no glitches). As I say that, one of my quads doing the sieving just re-booted all by itself.[/quote]


If you're in by Jan. 20th, we're good. Thanks for helping! :smile:

mdettweiler 2009-01-04 22:08

1 Attachment(s)
5850G-5900G is complete; factors are attached. :smile:

gd_barnes 2009-01-05 02:18

[quote=MyDogBuster;156674]It'll be tight for me but I think I'll sneak in just at the wire (assuming no glitches). As I say that, one of my quads doing the sieving just re-booted all by itself.[/quote]


Just a note: I am still assuming that you will be done with the P=1.4T-1.8T range here in the next day or two, which is what I think you had previously said. Is that correct?

I'll get the n=50K-200K portion of the 9th drive started when you're done.


Thanks,
Gary

MyDogBuster 2009-01-05 03:15

[QUOTE]Just a note: I am still assuming that you will be done with the P=1.4T-1.8T range here in the next day or two, which is what I think you had previously said. Is that correct?
[/QUOTE]

Still on schedule. Sometime tomorrow evening.

MyDogBuster 2009-01-05 21:16

1400G - 1800G complete

results emailed

gd_barnes 2009-01-06 02:37

I have now removed P=1.4T-1.8T factors from the file link in the 1st post of this thread. 80000+ factors total.

My sieving for P=1.8T-3.4T should be done by morning so I'll have a lot more factors to remove at that point. :smile:


Gary

Mini-Geek 2009-01-06 12:51

[quote=gd_barnes;157107]I have now removed P=1.4T-1.8T factors from the file link in the 1st post of this thread. 80000+ factors total.[/quote]
Would it be a problem in processing my factors if I drop that new file into my sieving to speed it up? Also I recall that the file can be dropped in over the old one without having to restart sr2sieve, is that right or am I mistaken?

gd_barnes 2009-01-06 13:32

[quote=Mini-Geek;157166]Would it be a problem in processing my factors if I drop that new file into my sieving to speed it up? Also I recall that the file can be dropped in over the old one without having to restart sr2sieve, is that right or am I mistaken?[/quote]


On the first question, no problem at all.

On the second question, I'm not sure. I think Max knows. But it's no problem to stop your sieve using CTL-C because it will write a checkpoint file. Then after dropping the file in, you can restart it and it will start right where it left off.

But hold off a couple of hours. I think my sieving has just now completed the P=1.8T-3.4T range. I need to remotely check my machines. If so, I'll remove those factors too and post an even smaller file sieved to P=3.8T since Chris has also finsihed P=3.4T-3.8T.

BTW, your sieving speed will likely increase very little...I'm thinking < 3% even with mine and Chris's factors removed...just don't want you to be disappointed.


Gary

gd_barnes 2009-01-06 15:21

P=1800G-3400G is complete.

Now reserving the (presumably) final range of P=8800G-10000G for work on 2 quads.

Mini and all, I've now removed all factors up to P=3800G from the linked sieve file in the 1st post of this thread. Feel free to replace your file with the file now there to speed up sieving a little bit.

Right on track so far... :smile:


Gary

gd_barnes 2009-01-06 15:42

Moved posts related to the discussion of how to do primality testing on the n=50K-200K range to the new 9th drive thread.

mdettweiler 2009-01-06 16:09

[quote=gd_barnes;157172]On the first question, no problem at all.

On the second question, I'm not sure. I think Max knows. But it's no problem to stop your sieve using CTL-C because it will write a checkpoint file. Then after dropping the file in, you can restart it and it will start right where it left off.[/quote]
No, you have to restart sr2sieve in order for the change to take effect, since it loads the sieve file into memory when it first starts. But, as Gary said, if you stop it "nicely" with Ctrl-C, then it will pick up exactly where it left off when you restart it.

Mini-Geek 2009-01-06 16:52

[quote=mdettweiler;157193]No, you have to restart sr2sieve in order for the change to take effect, since it loads the sieve file into memory when it first starts. But, as Gary said, if you stop it "nicely" with Ctrl-C, then it will pick up exactly where it left off when you restart it.[/quote]
Ok. I stopped it with Ctrl+C and restarted it with the new files already.

MyDogBuster 2009-01-07 17:11

3800G - 4000G Complete

results emailed

nuggetprime 2009-01-10 09:58

8200G-8600G complete.9616 factors found. They will get mailed to Gary with in the next fifteen minutes.

nugget

Mini-Geek 2009-01-10 12:32

1 Attachment(s)
7100G-7200G complete. 2829 factors attached.

glennpat 2009-01-12 22:25

1 Attachment(s)
8600G-8800G complete. 4655 factors attached.

gd_barnes 2009-01-12 23:31

[quote=nuggetprime;157884]8200G-8600G complete.9616 factors found. They will get mailed to Gary with in the next fifteen minutes.

nugget[/quote]


Nugget,

I haven't received your factors. You can go ahead and post them here if you want.


Thanks,
Gary

gd_barnes 2009-01-12 23:43

Ian and David,

Can you give me an ETA on your sieving ranges?

My ETA:
5900G-7100G; late on Jan. 16th running on 1 quad
8800G-10000G; mid-day on Jan. 13th running on 2 quads

If I seem in a bit of a hurry, I am. :smile:

PrimeGrid and others are dumping so many n=400K-500K primes on top-5000 that n=352K is already in ~4950th place on top-5000. If we get started by Jan. 17th, we'll probably be fine but if we get started by Jan. 20th, n=352K may already be below 5000th place.


Gary

IronBits 2009-01-13 00:38

Done, and emailed to you.

MyDogBuster 2009-01-13 00:49

4.25T - 5.85T - Late the 15th

4.05T - 4.25T - Real late the 16th

gd_barnes 2009-01-13 04:29

[quote=IronBits;158340]Done, and emailed to you.[/quote]


I don't see them in my Email. When did you send them? If I've inadvertantly deleted any Email in the last week, it's still in my trash and should be able to find it. I just now checked my trash and don't see it. I also checked my spam box and don't see it there either.

Has my Email gone bonkers? lol I'm still receiving plenty of Email from people; just not the factors files from David and Nugget here.


Gary

IronBits 2009-01-13 05:15

My BAD! I emailed them to Max because he sent me the work.
I have just now forwarded it to you.
Wasn't sure exactly what you would need, so grabbed it all
[code]
101 checkpoint-b.txt
101 checkpoint-c.txt
101 checkpoint-d.txt
101 checkpoint-e.txt
101 checkpoint-f.txt
101 checkpoint-g.txt
101 checkpoint-h.txt
230,175 factors-a.txt
98,439 factors-b.txt
95,535 factors-c.txt
95,898 factors-d.txt
94,314 factors-e.txt
91,278 factors-f.txt
90,948 factors-g.txt
86,658 factors-h.txt
19,089 sr2sieve-a.log
1,257 sr2sieve-b.log
1,257 sr2sieve-c.log
1,257 sr2sieve-d.log
1,257 sr2sieve-e.log
1,257 sr2sieve-f.log
1,256 sr2sieve-g.log
1,256 sr2sieve-h.log
[/code]

gd_barnes 2009-01-13 05:53

David,

Why did you say you were "Done and emailed to you" when your range is not done yet? (lol) Are you still sieving? I hope so.

Anyway, since you aren't done, all that I needed was an ETA (estimated completed date) not all of the files.

I went ahead and checked the files. The factors-a range of P=7200G to 7325G is done. The factors-b thru factors-h files are 85% done. You can now run the factors-a core on something else if you haven't already started.

---you can skip this fluff stuff if you want and go to the last para. that has helpful info.---

Since you're 85% complete on those files and I see by looking in this thread that you probably started on Jan. 3rd, that means you've been running 9 days.

Therefore: 9 days / 85% = 10.6 days

So, you should have 1.6 days remaining or a completion date of late on Jan. 14th. That sounds good assuming that you're still sieving. :smile:

---end of fluff stuff

Helpful sieving stuff for you:

1. Sr2sieve will show the date/time that it expects completion at the command prompt as it is displaying factors while it is running. That's where you could get the ETA without having to mess with the files.

2. The checkpoint.txt file will give you a percent completion. It is updated every 5 mins. That is where I got your % completion on your factors-b thru factors-h files. Here's an example of the checkpoint-b.txt file:
pmin=7431645640543,factor_count=2983,cpu_secs=654679.724,[B]frac_done=0.853165[/B],elapsed_secs=653410.920

The "frac_done" shows the percentage of completion, i.e. .853165. Actually, if you look at all of the other stuff that it shows, they are quite self explanetory even if you've never sieved before. It's a very helpful little file that allows you to restart right where you left off if the machine shuts down.

I hope this helps.


Gary

mdettweiler 2009-01-13 06:03

[quote=IronBits;158401]My BAD! I emailed them to Max because he sent me the work.
I have just now forwarded it to you.
Wasn't sure exactly what you would need, so grabbed it all
[code]
101 checkpoint-b.txt
101 checkpoint-c.txt
101 checkpoint-d.txt
101 checkpoint-e.txt
101 checkpoint-f.txt
101 checkpoint-g.txt
101 checkpoint-h.txt
230,175 factors-a.txt
98,439 factors-b.txt
95,535 factors-c.txt
95,898 factors-d.txt
94,314 factors-e.txt
91,278 factors-f.txt
90,948 factors-g.txt
86,658 factors-h.txt
19,089 sr2sieve-a.log
1,257 sr2sieve-b.log
1,257 sr2sieve-c.log
1,257 sr2sieve-d.log
1,257 sr2sieve-e.log
1,257 sr2sieve-f.log
1,256 sr2sieve-g.log
1,256 sr2sieve-h.log
[/code][/quote]
Hmm...something doesn't look right there. There shouldn't be any checkpoint.txt files if the range is complete.

Okay...I've looked at the files you sent me via email, and it seems that instance A finished its range, but then you stopped the rest of them manually with Ctrl-C at that point. The rest of them appear to be a few minutes away from finishing; check your email for more details on this.

Gary, if you already removed the factors from the sieve file, no biggie; even though David will be re-sending you the [I]completed[/I] factors files for his range later on, it still won't hurt even if many of them have already been removed since srfile automatically deals with duplicate factors.

Max :smile:

Edit: Ah, I see Gary beat me to the punch on this. lol :smile:

gd_barnes 2009-01-13 06:11

Good, I'm glad I'm not the only one that thought "done" meant his sieving was done. lol

I'm only removing factors in fully contiguous ranges as they are completed so nothing above P=4050G has been removed yet.

I went ahead and marked P=7200G-7325G as complete.

IronBits 2009-01-13 06:11

Well
CRAP CRAP CRAP
It was no longer running when I got home, so assumed it had finished.
CRAP CRAP CRAP
I've restarted it
CRAP CRAP CRAP
;)

IronBits 2009-01-13 06:18

1 Attachment(s)
Here's what I see now.
I guess it did the first part, now it's doing something else... correct? :(

gd_barnes 2009-01-13 06:18

[quote=IronBits;158423]Well
CRAP CRAP CRAP
It was no longer running when I got home, so assumed it had finished.
CRAP CRAP CRAP
I've restarted it
CRAP CRAP CRAP
;)[/quote]


Well CRAP! lmao :smile: Seriously, it's no big deal.

Do you know what caused it to stop running? Did you have a power blip? Anyway, that's only a few hours loss. We're still in excellent shape as you should only have ~2 days left on them.

What happened to you is exactly why sr2sieve creates a checkpoint.txt file. If you restarted it in the exact same manner that you started it to begin with, then it should pick up right where it left off.

Just to verify: You still kept all of the files in the same folders (directories) that they were in previously. Is that correct?


Gary

IronBits 2009-01-13 06:25

Now it says 21 January, when last night it was saying 12 January at 10AM - WTF???
Of course I didn't touch anything...
hmmm, maybe I had one of them dyslexic moments.... /shrugs and walks away (I know, no smilies) ;) :wink:

mdettweiler 2009-01-13 06:29

[quote=IronBits;158429]Here's what I see now.
I guess it did the first part, now it's doing something else... correct? :([/quote]
Uh...it looks like you started instance A eight times, instead of running instances B-H. And since it's already done with instance A's range, it started from zero on all of those!

Here's what you need to do:
-Stop all your sr2sieve instances.
-Delete the file checkpoint-a.txt if it exists.
-Now, run the batch file start2.bat.
-Run start3.bat.
-Run start4.bat.
-Run start5.bat, etc. up to start8.bat.
(Note that you're *not* running start1.bat, since that's instance A which is the one instance that finished its work.)
-Wait for each of the instances to terminate, and *then* your range is complete. Zip up all the factors-[i][a through h][/i].txt files and send them to Gary.

That should do the trick. :smile:

Next time I need to send you a care package, if I label the instances with letters (a, b, c, etc.) I'll label the respective batch files with letters too, rather than with 1, 2, 3, 4, etc. I think that may have contributed to the confusion here. :smile:

Max :smile:

gd_barnes 2009-01-13 06:36

[quote=mdettweiler;158435]Uh...it looks like you started instance A eight times, instead of running instances B-H. And since it's already done with instance A's range, it started from zero on all of those!

Here's what you need to do:
-Stop all your sr2sieve instances.
-Delete the file checkpoint-a.txt if it exists.
-Now, run the batch file start2.bat.
-Run start3.bat.
-Run start4.bat.
-Run start5.bat, etc. up to start8.bat.
(Note that you're *not* running start1.bat, since that's instance A which is the one instance that finished its work.)
-Wait for each of the instances to terminate, and *then* your range is complete. Zip up all the factors-[I][a through h][/I].txt files and send them to Gary.

That should do the trick. :smile:

Next time I need to send you a care package, if I label the instances with letters (a, b, c, etc.) I'll label the respective batch files with letters too, rather than with 1, 2, 3, 4, etc. I think that may have contributed to the confusion here. :smile:

Max :smile:[/quote]


Max,

Is it possible that he deleted his checkpoint files from the various directories when he sent them to us? If so, would your process work?

One other thing: If his ranges were showing a completion of Jan. 12th on Sunday, since Monday is Jan. 12th, is there something wrong in your process for the ranges that he is sieving? In other words, did they THINK that they were only supposed to go as far as they did and then stopped?

David,

Why did your machines stop sieving without completing? Power blip? If they just stopped like that, then we may have a bigger problem.


Gary

IronBits 2009-01-13 06:38

:missingteeth:

Nope, I was a dumb A$$ - I had 8 instances of the line below...
start "sr2sieve-1" sr2sieve.exe -p 7200e9 -P 7325e9 -i sieve1005-2000-200K-500K.txt -u a -z
Basically, I put all 8 cores on A
Now I'm using this - can I put core A on B to help it finish faster?

start "sr2sieve-2" sr2sieve.exe -p 7200e9 -P 7325e9 -i sieve1005-2000-200K-500K.txt -u b -z
start "sr2sieve-3" sr2sieve.exe -p 7200e9 -P 7325e9 -i sieve1005-2000-200K-500K.txt -u c -z
start "sr2sieve-4" sr2sieve.exe -p 7200e9 -P 7325e9 -i sieve1005-2000-200K-500K.txt -u d -z
start "sr2sieve-5" sr2sieve.exe -p 7200e9 -P 7325e9 -i sieve1005-2000-200K-500K.txt -u e -z
start "sr2sieve-6" sr2sieve.exe -p 7200e9 -P 7325e9 -i sieve1005-2000-200K-500K.txt -u f -z
start "sr2sieve-7" sr2sieve.exe -p 7200e9 -P 7325e9 -i sieve1005-2000-200K-500K.txt -u g -z
start "sr2sieve-8" sr2sieve.exe -p 7200e9 -P 7325e9 -i sieve1005-2000-200K-500K.txt -u h -z

You fellers are way ahead of me, Max, I use start, in one .bat file. We are back on track.

mdettweiler 2009-01-13 06:41

[quote=IronBits;158442]:missingteeth:

Nope, I was a dumb A$$ - I had 8 instances of the line below...
start "sr2sieve-1" sr2sieve.exe -p 7200e9 -P 7325e9 -i sieve1005-2000-200K-500K.txt -u a -z
Basically, I put all 8 cores on A
Now I'm using this - can I put core A on B to help it finish faster?

start "sr2sieve-2" sr2sieve.exe -p 7200e9 -P 7325e9 -i sieve1005-2000-200K-500K.txt -u b -z
start "sr2sieve-3" sr2sieve.exe -p 7200e9 -P 7325e9 -i sieve1005-2000-200K-500K.txt -u c -z
start "sr2sieve-4" sr2sieve.exe -p 7200e9 -P 7325e9 -i sieve1005-2000-200K-500K.txt -u d -z
start "sr2sieve-5" sr2sieve.exe -p 7200e9 -P 7325e9 -i sieve1005-2000-200K-500K.txt -u e -z
start "sr2sieve-6" sr2sieve.exe -p 7200e9 -P 7325e9 -i sieve1005-2000-200K-500K.txt -u f -z
start "sr2sieve-7" sr2sieve.exe -p 7200e9 -P 7325e9 -i sieve1005-2000-200K-500K.txt -u g -z
start "sr2sieve-8" sr2sieve.exe -p 7200e9 -P 7325e9 -i sieve1005-2000-200K-500K.txt -u h -z[/quote]
This still won't work--check your email for full details.

gd_barnes 2009-01-13 06:52

I'll step out of this conversation...I could only hurt things. I'm not familiar with the multi-threading or whatever it is.

David, the "-p 7200e9 -P 7325e9" part of the command line is different for each instance.


Gary

IronBits 2009-01-13 06:55

I won't make that mistake again :cry:

Lennart 2009-01-13 07:02

[quote=IronBits;158449]I won't make that mistake again :cry:[/quote]


:missingteeth: That one was good :smile:

/Lennart

gd_barnes 2009-01-13 07:22

Sorry...I gotta step back in here. Max and David, the latest restart of sr2sieve is still not right. See your Emails. The ranges should take no more than 2 days to complete if restarted correctly.

Max, it has started each range over from the beginning. That is why it's showing an ETA of Jan. 20th, which is too late for our needs.

Gary

gd_barnes 2009-01-13 09:37

David and I talked through the sieving issues on the phone. Everything is good now. His ETA for completing his range is now early on Jan. 14th. :smile:


All times are UTC. The time now is 07:12.

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2021, Jelsoft Enterprises Ltd.