Author: Tord Romstad
Date: 02:40:12 05/28/05
Go up one level in this thread
On May 27, 2005 at 11:04:02, Vincent Diepeveen wrote: >On May 27, 2005 at 10:37:50, Tord Romstad wrote: > >>On May 27, 2005 at 08:28:12, Vincent Diepeveen wrote: >>> >>>How much money are you willing to bet on this indoctrination attempt by the >>>GIMPS guys? > >>What indoctrination attempt are you talking about? I provided a link to >>the GIMPS site in order to inform about the hardware and software used to >>find the currently biggest known prime number, and also because it contains >>a number of useful links for those who are interested in learning more >>about Mersenne primes and prime numbers in general. > >"world record primes are always Mersenne primes" > >that's a very rude statement and just wasn't true in the past and most likely >won't be true in the future either. Point taken -- I should be careful to use strong words like "always" without adding qualifications. In the quoted sentence above, it would definitely have been better to write "almost always" or to add the word "recently" somewhere. If I recall correctly (and it is very possible that I don't) a non-Mersenne number has been the biggest known prime only once in the computer era. It could happen again, of course, though I think Mersenne primes will still top the list most of the time. At any rate, I apologize for the inaccuracy. However, I still don't see what this has to do with GIMPS. They can just drop the letter 'M' from their name and search for non-Mersenne primes as well if they prefer. >the GIMPS project has what is it, a record of 5 or 6 primes now at their conto? I have no idea. >This because they hardly have any competition. Yeah. Prime hunting will never be among the world's biggest and most popular sports, I guess. :-) >>I couldn't care less who or what finds the first 10 or 100 million digit >>prime number. World record prime numbers in themselves hold absolutely >>zero interest for me. > >I am not careless about such huge achievements. Like most people, I tend not to care much about huge achievements in fields in which I don't have much interest. For instance, becoming an Olympic champion in badminton is arguably a much bigger achievement than finding a world record prime number (because the competition is much harder), but I still don't care who wins in the next Olympic games. >>Actually spending thousands of CPU hours hunting for new world >>record prime numbers seems like a tremendous waste of time to me. > >Please tell me, what do you consider useful ways to spend system time? Easy: Whatever ways you or your employer considers useful. This is an entirely subjective question. I often spends lots of CPU time running matches between computer chess programs, which almost everybody on the planet except those reading this forum would consider a waste of time. If you meant to ask me about examples of tasks similar to prime hunting which I would consider more interesting (and again, this is nothing more than a subjective opinion), I would suggest searching for an odd perfect number as a good example. After all, we know that there are infinitely many prime numbers, but we are still not sure whether there exists odd perfect numbers. From a mathematical perspective, finding a concrete prime number bigger than all those previously known is rather dull and unimportant, while finding an odd perfect number would be a surprising and very important number-theoretic discovery. >Note that research after new medicines and other bio-medical researches eats >less than 0.5% system time. > >I remember a research, trying to predict the future height of the ocean, running >on a big 1024 processor supercomputer here in Netherlands. After putting in half >a million hours of cpu time (yes that was many months of non stop computing at >hundreds of processors), they concluded that the sea would rise 1 meter. > >1 month later i found in a logfile of them: "small initialisation bug, we >initialised the ocean water height 1 meter too high". > >But was it wasted system time? > >The researchers got their publicity world wide and their own discovery show that >the ocean floor 'soon' will rise 1 meter. Cool story! :-) >>>Factorization goes faster than proving, >> >>I am sure you meant to say the opposite, as is evident from the rest of >>your post. Testing whether a big number is prime is much less work than >>actually factoring the number (which is also intuitively obvious). > >Please reread what i wrote. It's far easier to get rid of a big % of composites >by first factorising. To prove 1 number you'll have to do a million or 33+ >million FFT calculations. Sorry, it wasn't clear to me what you meant. It now seems that that you only meant looking for one small factor of a huge number, not about finding a complete factorization. In this case I agree with you, of course. Tord
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.