I return empty handed for these:
3321931433
3321931513
3321931531
3321931619

It seems that these two [url=http://2721.hddkillers.com/compare/live/]have been missed[/url] on the Eleven Smooth site:
[url]http://www.mersenneforum.org/showthread.php?p=192508#postcount192508[/url]

Incidentally, is there any particular reason why the bit level for 3321931061 is listed as "75?"?

[QUOTE=lavalamp;212389]I return empty handed for these:[/QUOTE]Did you finish the planned work on them?

Yes, searched them all to 76 bits without finding any factors.

[QUOTE=lavalamp;212397]Incidentally, is there any particular reason why the bit level for 3321931061 is listed as "75?"?[/QUOTE]

Post #335 asks if you finished to level 75 after finding the factor.
Post #336 says yes, to 75

Looks like I missed the #336 - I've change the "75?" to "75"

[QUOTE=lavalamp;212397]It seems that these two [url=http://2721.hddkillers.com/compare/live/]have been missed[/url] on the[/QUOTE]I just ran some calculations and I get an average bit depth of 75.92 via one set of calculations and 75.90 via a different set.
How are calculating it?

The 75.92 matches this: [url]http://www.moregimps.it/billion/expos_stats.php[/url]

I calculate it in a way that respects the extra work required to reach the higher bit levels:
log_2( (48 * 2^75 + 35 * 2^76 + 12 * 2^77 + 12 * 2^78 + 2^79) / 108 ) = 76.364054

So if all candidates were sieved equally deep, the bit level would be at 76.364054. That seems to make more sense as an average to me.

Huge reservation
 

I am taking ALL factored exponents to 2[sup]71[/sup] :smile:

I am also testing TheJudger's mfaktc on all OBD numbers to double-check them up to 2[sup]71[/sup].

I'll let you know about the checking.

Luigi

[QUOTE=ET_;214046]I am also testing TheJudger's mfaktc on all OBD numbers to double-check them up to 2[sup]71[/sup].[/QUOTE]
*cough*
[url]http://v5www.mersenne.org/report_factoring_effort/?exp_lo=332192831&exp_hi=332299999&bits_lo=0&bits_hi=70&txt=1&exassigned=1&B1=Get+Data[/url]
:bow:

[QUOTE=Uncwilly;214055]*cough*
[url]http://v5www.mersenne.org/report_factoring_effort/?exp_lo=332192831&exp_hi=332299999&bits_lo=0&bits_hi=70&txt=1&exassigned=1&B1=Get+Data[/url]
:bow:[/QUOTE]

???

You want ET_ to test 100M digit candidates? Why?

[QUOTE=axn;214058]???

You want ET_ to test 100M digit candidates? Why?[/QUOTE]

It could be a good idea to test mfaktc. The software already correctly finds 100MDPP factors but a double (or triple) check may be helpful.

Note that checking one OBD exponent from 1 to 69 requires about 90 [COLOR="Red"]minutes[/COLOR] with Factor5 (2 threads) and less than 85 [COLOR="Red"]seconds[/COLOR] with mfaktc.

Faster than Prime95...

Note, however, that 100MDPP numbers are about 3-4 times harder than OBD at the same bit-length; It may take more than 4 days to test ALL 870 exponents ...

I may retest groups of 89/90 exponents at a time, not more, due to restrictions to the use of my PC (the PC is barely usable when GPU is working at full speed, and I have other projects running).

Luigi