Just noticed we now have 3 100M completions.
 

...and the latest was error free.

[url]http://www.mersenne.org/report_LL/?exp_lo=100000000&exp_hi=999999999&exp_date=&user_only=0&user_id=jg5&dispdate=1&B1=Get+LL+data[/url]

[QUOTE=petrw1;305085]...and the latest was error free.

[url]http://www.mersenne.org/report_LL/?exp_lo=100000000&exp_hi=999999999&exp_date=&user_only=0&user_id=jg5&dispdate=1&B1=Get+LL+data[/url][/QUOTE]

Congratulations to the successful tester! This is an event that is a significant one in GIMPS history, and deserves mention on the milestones page, if not on the GIMPS home page itself. Way to go!

That's what a workstation with ECC can do. Good job!

That motherboard doesn't support ECC memory, but still, nice result!!

bcp19, what's your ETA?

[QUOTE=Dubslow;305115]That motherboard doesn't support ECC memory, but still, nice result!!
[/QUOTE]
Incorrect. [URL="http://www.asus.com/Motherboards/Intel_Socket_1366/P6T7_WS_SuperComputer/"]P6T7_WS[/URL].

I've looked it up way back when this reservation was already mulled over in the sewing circles.

Congratulations!
Now, who is going to be the first to double-check? LOL

I think this (doublecheck) does need to be p95, doesn't it? Or did the newer versions of CudaLucas over-passed the old FFT size limitation? (for this expo, about 20M FFT would be need, if I am not wrong, and CL was limited to 8M, then later to 16M). I just tried "-cufftbench" from 6 to 20 megs with a 65k step and it crased before 16M.

[QUOTE=LaurV;305183]I think this (doublecheck) does need to be p95, doesn't it? Or did the newer versions of CudaLucas over-passed the old FFT size limitation? (for this expo, about 20M FFT would be need, if I am not wrong, and CL was limited to 8M, then later to 16M). I just tried "-cufftbench" from 6 to 20 megs with a 65k step and it crased before 16M.[/QUOTE]
I've previously had problems with [URL="http://mersenneforum.org/showthread.php?p=302496#post302496"]cufftbench crashing[/URL]; I wouldn't take it to mean that CUDALucas can't use those FFT lengths.

You should be able to use up to a 64M FFT length, though you may need to increase threads [URL="http://mersenneforum.org/showthread.php?p=295750&highlight=brain#post295750"]passed 256[/URL]. (The error that Brain encountered now has a second line giving useful instructions on what to do in such a case.)

Here is the '[STRIKE]monthly[/STRIKE]' status report for the range from 332192831 to 332399999 (plus a bit more):

[code]Date of data 7/20/2012
Average bit depth for first 100 expos 78.41
Average bit depth for first 1000 expos 77.16
Average bit depth for first 5000 expos 73.40
Average bit depth for first 10000 expos 71.79
100th active expo (no factor found) 332198357
1000th active expo (no factor found) 332246143
5000th active expo (no factor found) 332438779
10000th active expo (no factor found) 332671049
Unitless total effort number 175,308,800
Number of first 100 expos to 2^77 89
Number of first 1000 expos to 2^76 885
Number of first 5000 expos to 2^77 774
Number of first 10000 expos to 2^76 1381
Number left in range 4205

Estimated expos in range to be removed 276
(by taking all expos to 2^79)[/code]
[code]Bit # at bit level
71 698
72 1369
73 335
74 377
75 131
76 533
77 298
78 30
79 404
80 25
81 5
P-1 201[/code]
LaurV has been churning through the range 332000000 - 332399999 elevating all below 69 to 70.
The attached show the % of prime exponents that are factored. The yellow arrow shows the effect of GPUto72, the purple arrow shows the effect of the 100M Digit Prefactor Project.

I finished the 332M-333M range to 70. This happened 2-3 days ago, but I first waited for the [URL="http://www.mersenne.info/trial_factored_tabular_data/3/332000000/"]visualization page[/URL] to catch up (and [URL="http://www.mersenne.info/trial_factored_bar_graph/3/332000000/"]make everything purple[/URL], as [URL="http://www.mersenneforum.org/showpost.php?p=303514&postcount=1377"]promised[/URL] :smile:, and as chalsall said, this is now not anymore so frequent as used to be, to protect the PrimeNet's pipelines), and then I forgot to report. On the way I found about 175 factors, which were saved in a file if someone needs them. Of course, an equivalent amount of exponents were cleared. :razz:

Edit: now because [URL="http://www.mersenneforum.org/showpost.php?p=306165&postcount=933"]me pillaged enough P-1 thingies[/URL], me is going back to [URL="http://www.mersenneforum.org/showthread.php?p=304280#post304280"]pay these two landlubbers[/URL], yaaarrrrr!!!.

Reserving 334M (all) to 66. This batch will take me about a week. Potentially taking to 67 next month. IK its not technically in this subproject, but I imagine there would be a several mil expansion once TF is done to an appropriate level for all y'all's range...

PS "all y'all's" is as close to Proper Hick as I can get.
Layers of y'all
Y'all: you
all y'all: all of y'all
All Y'all's: Belonging to all y'all (see above)