![]() |
![]() |
#1 |
Apr 2018
USA
13 Posts |
![]()
Will prime95/mprime continue storing results forevert, if it is unable to connect to the internet. I have surmised that copying certain files from the working directory on a machine running mprime/prime95 that is not connected to a network, to a machine that is connected, long enough to upload results and acquire more work, on a periodic basis, might remedy the problem of no network access.
But is there some method that is preferred to avoid pitfalls that may not be obvious at first. Would I simply specify to obtain 6 months of work, and then upload results every 4th month. And will my idea even work at all? Using Linux. Thanks for all the help. |
![]() |
![]() |
![]() |
#2 |
Oct 2020
Terre Haute, IN
23×13 Posts |
![]()
It might be frowned upon to try to reserve multiple months worth of work because it might appear you are "hoarding" exponents. There are far too many exponents already which have been reserved and no work has been done on them in several years. Here's one I noticed hadn't had any apparent progress on it in more than six years:
https://www.mersenne.ca/exponent/332197973 One month worth of work at a time might be much more preferable. |
![]() |
![]() |
![]() |
#3 |
"TF79LL86GIMPS96gpu17"
Mar 2017
US midwest
3×5×431 Posts |
![]()
Results files are as permanent as the storage device allows. Assignments expire. Read the assignment rules. Only take on work that is reserved by you, and won't expire before you will report results for it. If queuing up a machine deeply in an isolated state, make sure it's work that won't get in the way of making progress at the wavefronts of first time testing, P-1 in preparation, double-checking, etc. Think high Cat 4 assignments. PRP with GEC and proof generation will mean large proof files to be transferred back. It's likely few if any would mind if you did mass 100Mdigit PRP/GEC/proof as DC of some existing unverified 100Mdigit LL tests. There are over 840 currently. https://www.mersenne.org/report_ll/?...dispdate=1&B1= Those are 2 weeks each on a Radeon VII GPU, and several weeks or months on most CPUs. And just don't do LL first tests. They are now a complete waste of time.
Last fiddled with by kriesel on 2021-10-08 at 00:59 |
![]() |
![]() |
![]() |
#4 |
Apr 2018
USA
1310 Posts |
![]()
I've already started work. I just chose the default type of work. I moved the working directory onto the isolated machine, and it's been going for about as long as this post has been up. But I bookmarked this post for future reference. I've been on Primenet since about 2003. Yet, I haven't taken much time to learn about it.
I got into it by googling for a cpu benchmark/burn-in program for my Pentium 4 machines. I was taken not by the mathematics--although I'm an engineer--nor the behemoth nature of the task at hand, but by the scarcity of Mersenne primes. And, I must admit the concept of large primes was--and is--intriguing. I'm not in the top 500 any longer, so I don't want a single possible cpu cycle wasted. Hence, the isolated machine is employed for the GIMPS. Now for the question of the day, "Could anyone possibly care less about my little anecdote?" ![]() Last fiddled with by AwesomeMachine on 2021-11-05 at 04:17 Reason: addition |
![]() |
![]() |
![]() |
Thread Tools | |
![]() |
||||
Thread | Thread Starter | Forum | Replies | Last Post |
Add option to submit unsuccesful ECM work done | bur | FactorDB | 8 | 2021-05-21 12:39 |
Automatic submit results + fetch assignments for mfaktc? | DuskFalls | GPU Computing | 5 | 2017-12-02 00:34 |
How do we prevent miraculously true DC results from manual submit? | leonardyan96 | PrimeNet | 77 | 2017-06-01 16:18 |
MLucas, submit results? | Sleeping_menace | Mlucas | 17 | 2015-06-13 03:12 |
Only submit part of ECM results? | dabaichi | PrimeNet | 5 | 2011-12-07 19:27 |