Riesel base 3 reservations/statuses/primes

[rogue]

Originally Posted by KEP

Yes it will, because as soon as my R3 to n=100K (8G-13G) is complete, I'm going to put 8 cores on the effort if Gary feels reliable with srbsieve and can approve of the use of srbsieve. Now I just hope, that I can at least be allowed to reserve a 10G range, to n=25K, since it would make it easier to prepare the workfiles and completion will be in about 4-5 weeks... so please Gary, could you, if testing of srbsieve turns out as we all hope, accept a 10G reservation for me?

Guess it is now also time to consider starting up S3

Tomorrow, unless lalera is working on this, I'll use the starting bases script to test k=20,000,000,002 to k=20,100,000,000 to n=24 and then copy paste the primes with n>24 from the pl_primes.txt file created by srbsieve, and then see if the .txt files match each other. Hopefully within the weekend we will know if both ranges match ... can this btw be accepted as a correct way to test the reliability of srbsieve or what do you think Gary? Rogue?

With srbsieve you delete all of the phase lines. It will then stop after filtering the newpgen files.

You will have to use the second version I posted as it will output pl_remain.txt correctly (the first version was missing *b^n+c" on each line.

It might be worth considering running srbsieve to n=100 (with two more phases) and compare with new-base.txt to n=100 and verify those as well. Gary makes the final call on what is acceptable.

[lalera]

Originally Posted by KEP

Yes it will, because as soon as my R3 to n=100K (8G-13G) is complete, I'm going to put 8 cores on the effort if Gary feels reliable with srbsieve and can approve of the use of srbsieve. Now I just hope, that I can at least be allowed to reserve a 10G range, to n=25K, since it would make it easier to prepare the workfiles and completion will be in about 4-5 weeks... so please Gary, could you, if testing of srbsieve turns out as we all hope, accept a 10G reservation for me?

Guess it is now also time to consider starting up S3

Tomorrow, unless lalera is working on this, I'll use the starting bases script to test k=20,000,000,002 to k=20,100,000,000 to n=24 and then copy paste the primes with n>24 from the pl_primes.txt file created by srbsieve, and then see if the .txt files match each other. Hopefully within the weekend we will know if both ranges match ... can this btw be accepted as a correct way to test the reliability of srbsieve or what do you think Gary? Rogue?

this
"i can do a range of 100M - 20,000,000,002 to 20,100,000,000 if you like"
was meant only as an offer (as a range of 1G is too big for me)

you reserved about 1 hour later this range officially so i decided not to do this range

[KEP]

Originally Posted by lalera

this
"i can do a range of 100M - 20,000,000,002 to 20,100,000,000 if you like"
was meant only as an offer (as a range of 1G is too big for me)

you reserved about 1 hour later this range officially so i decided not to do this range

Thanks, I just had to be sure, there is no reason to double our work, there is plenty of work as it is, so we all have to work on our own seperate reservations. I understand, that 1G can be to much, especially as a testing range, but if you do want to start up new bases like base 3, 7 and 15, it is the most leisure to reserve a 1G range at a time.

Again thanks for your feedback

[KEP]

Originally Posted by rogue

With srbsieve you delete all of the phase lines. It will then stop after filtering the newpgen files.

You will have to use the second version I posted as it will output pl_remain.txt correctly (the first version was missing *b^n+c" on each line.

It might be worth considering running srbsieve to n=100 (with two more phases) and compare with new-base.txt to n=100 and verify those as well. Gary makes the final call on what is acceptable.

I'll see what Gary accepts/suggests. But for the second testing, I'm gonna do a test with several phases of srbsieve (latest version) and I'm gonna run the same range of k's and n's using the starting bases script, then it will all be a matter of sorting and comparing

Looking forward to be hearing what Gary comes back with, but I might consider running to n=300 and thereby include phase 1 and phase 2 in srbsieve after removing newpgen found primes for n<=24. Unless Gary would like otherwise, I do believe a test to n=300 is sufficient, since I have now almost completed a flawless test through 6 phases to n=25K for version 1 of srbsieve, so since it is for comparison only, I think n=300 makes perfect sence. Gary?

[KEP]

Officially reserving:

k=20,100,000,002 to k=23,000,000,000 for my Hasswell. Work has been prepared for this range.

I'm gonna use srbsieve Version 4 and PFGW version 3.7.9

Completed using srbsieve Version 2 and PFGW version 3.7.9 is the following range:

k=20,000,000,002 to k=20,100,000,000 with following stats:

35,440,051 primes
14,559,376 MOB's
573 remains

50,000,000 Total

Result files will be sent as soon as upload to my Google drive completes

[VBCurtis]

Taking 63G and up to test srbsieve for myself. I'll run to at least 50k.

[VBCurtis]

The test was indeed quick (at 43k now, still running with just 50 k's left). I'm ready to tackle a 1G range to 25k. What file-transfer services are the preferred standards these days? How large a chunk can I srbsieve at once without making files too big to transfer? I already built the newpgen files for 62-63G, which are enormous! (over 15GB).

Taking 62-63G.

Also, is there a resume feature at some stage of srbsieve? i'm nervous my laptop won't finish a 250 or 500M chunk of srbsieve without a restart.

[rebirther]

Reserving R3 from n=100-150k (4-5G) for BOINC

[KEP]

Originally Posted by VBCurtis

The test was indeed quick (at 43k now, still running with just 50 k's left). I'm ready to tackle a 1G range to 25k. What file-transfer services are the preferred standards these days? How large a chunk can I srbsieve at once without making files too big to transfer? I already built the newpgen files for 62-63G, which are enormous! (over 15GB).

Taking 62-63G.

Also, is there a resume feature at some stage of srbsieve? i'm nervous my laptop won't finish a 250 or 500M chunk of srbsieve without a restart.

I'm not sure at what speed your laptop can complete a 1G range. However since there is no resume function in srbsieve, I suggest you do following approach:

Sieve k=62,000,000,001 to k=63,000,000,000 to n=XX using NewPGen, you can sieve the entire 1G range at once, since it won't slow srbsieve down, compared to the time saved on sieving.

After sieving, prepare 10 ranges, or what ever you are comfortable running and then start running the ranges 1 by 1 untill you complete all of them.

If you compress the pl_MOB.txt, pl_remain.txt and pl_prime.txt files, produced by srbsieve, using WinRAR then your compressed .rar file will be approximately 800MB to 1GB. I use GoogleDrive to send the compressed files to Gary, but other people may have other suggestions, however GoogleDrive is very reliable and transfers as fast as Gary can download the files sent to him.

Hope this helped you and welcome aboard the effort to bring R3 to it's knees. I'll continue move from the bottom and up, but will you continue towards the bottom from the top? If you do, then we can meet like the people digging a tunnel in 3-5 months time and declare R3 completely tested to n>=25K

Take care

[KEP]

Range 20,000,000,002 to 23,000,000,000 is close to complete. (Saturday or sunday)

I'm currently preparing work for k=23,000,000,002 to k=27,000,000,000 to n=25K, so therefor I'm officially reserving k=23G-27G.

[gd_barnes]

Originally Posted by VBCurtis

The test was indeed quick (at 43k now, still running with just 50 k's left). I'm ready to tackle a 1G range to 25k. What file-transfer services are the preferred standards these days? How large a chunk can I srbsieve at once without making files too big to transfer? I already built the newpgen files for 62-63G, which are enormous! (over 15GB).

Taking 62-63G.

Also, is there a resume feature at some stage of srbsieve? i'm nervous my laptop won't finish a 250 or 500M chunk of srbsieve without a restart.

All that I need are the primes and k's remaining.

Here is what works easiest for me for bases 3/7/15 for very large k-ranges:
1. Test to n=25K before sending anything. (I assume that you will be doing this.)
2. Send me all the k's remaining in one file. It won't be too big.
3. Divide the primes up into approximately k=100M or 200M files and zip them.
4. Upload all the zip files to google drive.
5. Send me a link to the google drive file.

If you want, you can skip uploading the k's remaining to google drive and just send me the file by itself.