Oops - New Prime! (M49 related) - Page 48

VBCurtis · 2016-01-22, 19:25

I concur- my first assignments in 1997 took about 25 days on a Celeron300 @ 504mhz. The fastest P3 of the era was 500Mhz.

I found Anandtech, began overclocking, and found GIMPS shortly afterward.

Spherical Cow · 2016-01-22, 19:42

It seems I remember them taking much longer. I think I ran Prime95 in the very early days, was absent for a while, and then restarted in 2002 or so. At that time, a single LL on my machine took substantially more than a month (I think); now I'm turning in two a month. My impression would be that machine speed-up is outpacing the slowdown due to exponent size.

Strictly from memory, though...And it may be my memory that's old and slow.

Norm

Zeta-Flux · 2016-01-22, 20:25

A belated congratulations to all involved!

By the way, is it possible that there are similarly unaccounted for primes in the 17 or bust project, collecting dust somewhere? It has just been way too long since they found their last prime.

ET_ · 2016-01-22, 20:57

Quote:

Originally Posted by petrw1

Seems YES: My first Pentium 4 from about 22 years ago took 1 month for a LL test in the current Leading Edge 33M range.

In fact, I had to wait one month for my PentiumPro 200 MHz to complete 1,500,xxx.

Mark Rose · 2016-01-22, 20:58

Quote:

Originally Posted by Zeta-Flux

A belated congratulations to all involved!

By the way, is it possible that there are similarly unaccounted for primes in the 17 or bust project, collecting dust somewhere? It has just been way too long since they found their last prime.

That's where I dedicate my CPU resources. SoB doesn't double check immediately, so it's possible a prime was missed.

There are only 100-150 CPU full-time modern cores equivalent currently working at SoB, so progress is slow. There is some overlap with the Prime Sierpinski Project, so that doesn't include all the effort towards finding the remaining 6 primes to be found. Also, there is other work being done via Boinc.

M29 · 2016-01-22, 21:30

The teraflops/second metric just spiked. What happened?

ixfd64 · 2016-01-22, 21:32

Quote:

Originally Posted by M29

The teraflops/second metric just spiked. What happened?

The influx of new users returning their double-check results?

petrw1 · 2016-01-22, 21:39

Quote:

Originally Posted by M29

The teraflops/second metric just spiked. What happened?

140,000 GhzDays of ANONYMOUS DCTF here:
http://www.gpu72.com/reports/worker_...b08af536d6adf/

VBCurtis · 2016-01-22, 22:20

Quote:

Originally Posted by Mark Rose

That's where I dedicate my CPU resources. SoB doesn't double check immediately, so it's possible a prime was missed.

There are only 100-150 CPU full-time modern cores equivalent currently working at SoB, so progress is slow. There is some overlap with the Prime Sierpinski Project, so that doesn't include all the effort towards finding the remaining 6 primes to be found. Also, there is other work being done via Boinc.

I believe the "similarly unaccounted" that you replied to had in mind a prime that a computer found but a human hadn't seen yet. I agree that SoB is somewhat likely to find a prime via doublecheck, and I doubt a prime is lying in a results file not yet noticed. The lack of record-ness of their size of prime means there is no attempt made to hide "Prime!" results upon return to the server, unlike GIMPS.

TObject · 2016-01-22, 22:25

Quote:

Curiously, Cooper says that the computer's notification system also failed the four other times it found a record number.

What kind of notification was expected? Sound?
Was it the NT Service flavor of Prime95? Services cannot make sounds.

ixfd64 · 2016-01-22, 22:28

Quote:

Originally Posted by VBCurtis

I believe the "similarly unaccounted" that you replied to had in mind a prime that a computer found but a human hadn't seen yet. I agree that SoB is somewhat likely to find a prime via doublecheck, and I doubt a prime is lying in a results file not yet noticed. The lack of record-ness of their size of prime means there is no attempt made to hide "Prime!" results upon return to the server, unlike GIMPS.

They actually did once find a new prime via double-checking.

2016-01-22, 21:30	#523
M29 Dec 2003 164₁₀ Posts	The teraflops/second metric just spiked. What happened? Attached Thumbnails

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Are Bitcoins Prime Related	a1call	Miscellaneous Math	23	2020-09-17 13:17
Merry Christmas and a prime! (M50 related)	Prime95	News	505	2020-01-18 01:03
Oops i did it again. (Prime found)	ltd	Prime Sierpinski Project	21	2006-01-04 14:50
Another new prime (M42-Related)	Uncwilly	News	132	2005-05-10 19:47
some prime-related trick questions	ixfd64	Puzzles	2	2003-09-23 12:53

2016-01-22, 19:25	#518
VBCurtis "Curtis" Feb 2005 Riverside, CA 1243₁₆ Posts	I concur- my first assignments in 1997 took about 25 days on a Celeron300 @ 504mhz. The fastest P3 of the era was 500Mhz. I found Anandtech, began overclocking, and found GIMPS shortly afterward.

2016-01-22, 19:42	#519
Spherical Cow Nov 2004 540₁₀ Posts	It seems I remember them taking much longer. I think I ran Prime95 in the very early days, was absent for a while, and then restarted in 2002 or so. At that time, a single LL on my machine took substantially more than a month (I think); now I'm turning in two a month. My impression would be that machine speed-up is outpacing the slowdown due to exponent size. Strictly from memory, though...And it may be my memory that's old and slow. Norm

2016-01-22, 20:25	#520
Zeta-Flux May 2003 7·13·17 Posts	A belated congratulations to all involved! By the way, is it possible that there are similarly unaccounted for primes in the 17 or bust project, collecting dust somewhere? It has just been way too long since they found their last prime.