Bases 501-1030 reservations/statuses/primes

[rebirther]

@rogue: Do you have a guide how much VRAM will be used for x sequences with the srsieve2cl?

[KEP]

Originally Posted by rogue

I figured out why this is happening. It will be fixed in then next release.

For now avoid using -W with srsieve2cl. By default it will use -W0 and -G1. It will use the CPU for testing until p > 1e6 then switch to the GPU.

Thats a puzzle, shouldnt it start at p=37G in stead of p=0?

Good you figured it out.

[rogue]

Originally Posted by KEP

Thats a puzzle, shouldnt it start at p=37G in stead of p=0?

If you didn't provide -p on the command line and it isn't in the input file, it will start at p=0. Did I misunderstand your question?

[rogue]

Originally Posted by rebirther

@rogue: Do you have a guide how much VRAM will be used for x sequences with the srsieve2cl?

The code does capture the number of bytes it allocates in GPU memory needed for kernel arguments. I could modify the code to print that out when using -H.

[KEP]

Originally Posted by rogue

If you didn't provide -p on the command line and it isn't in the input file, it will start at p=0. Did I misunderstand your question?

Maybe I wasn't clear, I'm having too many balls in the air at the moment and the building of my 2 new machines turned out due to a human error (made by me) to be far more than a walk in the park.

The question was, as you could see in my copy paste of CMD prompt window of what in fact happens, that I used this flag "-p 37684705039" and therefor that should be the starting value of p, but it appears that srsieve2cl still started at p=0, is that correct or did I missread the outputs that was printed on the screen? ... Does the use of -W x somehow interfere with the -p setting?

[rogue]

Originally Posted by KEP

Maybe I wasn't clear, I'm having too many balls in the air at the moment and the building of my 2 new machines turned out due to a human error (made by me) to be far more than a walk in the park.

The question was, as you could see in my copy paste of CMD prompt window of what in fact happens, that I used this flag "-p 37684705039" and therefor that should be the starting value of p, but it appears that srsieve2cl still started at p=0, is that correct or did I missread the outputs that was printed on the screen? ... Does the use of -W x somehow interfere with the -p setting?

The issue is with -W in srsieve2cl. If you remove -W and have a single GPU worker and you only see p=0, then there could be a problem.

[rebirther]

S504 tested to n=300k (250-300k)

nothing found, 3 remain

Results emailed - Base released

[rebirther]

Reserving S516 to n=600k (400-600k) for BOINC

Reserving S534 to n=600k (400-600k) for BOINC

Reserving S574 to n=600k (400-600k) for BOINC

Reserving S579 to n=600k (400-600k) for BOINC

Reserving S582 to n=600k (400-600k) for BOINC

Reserving S596 to n=600k (400-600k) for BOINC

[MisterBitcoin]

Originally Posted by MisterBitcoin

R712 is at n=726K with around 1 hour and 40 minutes runtime, its slowly processing. I will focus on this one more, extending reservation up to 800K.

I added all of results again, since i dont realy know when i posted an update on it.

Testing reached n=800K. No primes found, results attached. Releasing this base.

[rebirther]

Reserving S605 to n=600k (400-600k) for BOINC

Reserving S649 to n=600k (400-600k) for BOINC

Reserving S676 to n=600k (400-600k) for BOINC

[rebirther]

S520 tested to n=1M (700k-1M)

nothing found, 1 remain

Results emailed - Base released